Objective
Over 1.6 million Americans are affected by type 1 diabetes (T1D), including more than 1.4 million adults. T1D is associated with an increased risk of chronic complications including cardiovascular disease (CVD) and chronic kidney disease (CKD). Despite this, there are few dedicated studies testing approaches to reduce cardiorenal risk in this population. Recently, several drugs approved for the treatment of type 2 diabetes (T2D), including SGLT2 inhibitors (SGLT2is), GLP-1 receptor agonists (GLP-1 RAs) and the mineralocorticoid receptor antagonist (MRA) finerenone have been shown to reduce the risk for major adverse cardiovascular events, hospitalization for heart failure and adverse kidney outcomes in high-risk patients with T2D. The overarching goal of this research proposal is to provide evidence from existing trials in T2D that could support the development and further testing of these drugs to reduce the risk of adverse cardiorenal outcomes in patients with T1D.
Hypothesis: We hypothesize that measurement of islet autoantibodies (IAA) and c-peptide in biorepository samples from patients enrolled in T2D CVOTs will identify 5-10% of patients with adult-onset autoimmune diabetes. In subgroup analyses, patients with probable T1D in these and other trials will have a similar safety and efficacy profile compared to the general trial population, as well as propensity-matched controls.
To test this hypothesis, we will pursue 2 Specific Aims:
1) To measure IAA (GAD, IA2, ZNT8) and c-peptide in ~9400 patients from 4 T2D CVOTs representing 3 classes of medications: SGLT2is (sotagliflozin, ertugliflozin), GLP-1RAs (semaglutide) and MRAs (finerenone). These data will be used to determine the prevalence of IAA positivity among high-risk patients in these CVOTs and to perform subgroup analyses comparing the safety and efficacy profile among patients with probable T1D to those with typical T2D.
2) Using data from these analyses as well as other bioinformatic approaches, to classify patients in all CVOTs with publicly available data as to their likelihood having probable T1D and perform similar subgroup analyses in this broader patient population to increase the strength of any safety and efficacy signals.
Significance: Findings of similar safety and efficacy profiles in IAA positive patients will support further development of SGLT2is, GLP-1RAs and/or finerenone for CVD and CKD risk reduction in patients with T1D.
Innovation: This proposal leverages the considerable patient exposure and existing biorepositories for 3 classes of drugs with proven cardiorenal benefits to provide preliminary data on the safety and efficacy of these drugs in patients with probable T1D in a rapid and cost-efficient manner.
Background Rationale
Background/Rationale: Over 1.6 million Americans are affected by type 1 diabetes (T1D), including more than 1.4 million adults. In children, the diagnosis of T1D is often made following the acute development of symptoms. However, T1D can occur at any age. Indeed, half of all cases are diagnosed in adults, who often have a much slower course. As a result, as many as 40% of cases of adult-onset T1D are misdiagnosed as type 2 diabetes (T2D). Although these individuals progress more rapidly to requiring insulin for glycemic control, they may remain misclassified as T2D for years. In multiple studies across several populations, between 5-10% of patients diagnosed as T2D are positive for islet autoantibodies (IAA), typically GAD65.
T1D is associated with an increased risk of chronic complications including cardiovascular disease (CVD) and chronic kidney disease (CKD) which lead to an average loss of life expectancy of 11 years and over $16 billion in healthcare expenditures and lost income annually. The DCCT/EDIC study demonstrated that early, intensive glycemic control decreased the risk of microvascular complications, including chronic kidney disease, and was associated with a 40-57% reduction in CVD events after ~20 years, although the study was not designed nor was it powered as a CV outcome trial (CVOT). Despite the high risk of CVD and CKD in T1D, there have been few dedicated studies testing interventions, apart from glycemic control, to reduce risk for these adverse outcomes in this population. Consequently, the approach to CV risk reduction in T1D has largely been extrapolated from studies in individuals with T2D.
In contrast to T1D, there has been considerable progress in reducing CV risk in patients with T2D in the last decade. Over eight classes of drugs including the SGLT2 inhibitors (SGLT2is), GLP-1 receptor agonists (GLP-1 RAs) and the mineralocorticoid receptor antagonist (MRA) finerenone have been studied in robust CVOTs. We recently published a meta-analysis of CVOTs with SGLT2is (6 trials, 46,969 patients) demonstrating that this class decreased risk for major adverse cardiovascular events (MACE) by 10%, hospitalization for heart failure (HHF) by 32% and composite kidney outcomes by 38%. We also recently published a meta-analysis of GLP-1RA CVOTs (8 trials, 60,080 patients) which demonstrated that this class reduced MACE by 14%, all-cause mortality by 12%, HHF by 11% and a composite kidney outcome by 21%. Finally, in 2 recent studies in patients with T2D and CKD, the novel MRA finerenone significantly reduced a composite CVD endpoint (CVD mortality, myocardial infarction, stroke, or HHF) by 13-14% and a composite kidney endpoint by 13-18%. The mechanisms for these beneficial effects are not known and likely differ between the various classes, but they appear independent of glucose lowering. Although these trials excluded patients with known T1D, between 45-60% of patients were treated with insulin. Applying earlier estimates of IAA prevalence in T2D, it is likely that up to 10% patients in these trials were IAA positive. Several of these trials collected plasma and serum samples in biorepositories. Thus, it is possible to retrospectively identify IAA positive patients in these trials to determine, in subgroup analyses, whether there are any safety signals or early signals of efficacy in patients with probable T1D, before embarking on expensive, long-term, dedicated trials in patients with T1D.
Description of Project
Over 1.6 million Americans are affected by type 1 diabetes (T1D), including more than 1.4 million adults. T1D can occur at any age and, indeed, half of all cases are diagnosed in adults. Adults diagnosed with T1D often have a much slower course, leading to misdiagnosis as type 2 diabetes (T2D) in as many as 40% of cases. Between 5-10% of patients diagnosed with T2D are positive for islet autoantibodies (IAA) and may actually have T1D. T1D is associated with an increased risk of chronic complications including cardiovascular disease (CVD) and chronic kidney disease (CKD). Among patients with T1D of 20-30 years duration, 50-65% will die from cardiovascular complications, while CKD increases mortality 2-20-fold. Despite the high risk of CVD and CKD in T1D, there are few dedicated studies testing approaches to reduce risk in this population. As a consequence, apart from glycemic control, the approach to CVD and CKD risk reduction in T1D has largely been extrapolated from studies in individuals with T2D.
During the last 2 decades, a number of robust cardiovascular outcome trials (CVOTs) have demonstrated that drugs approved for the treatment of T2D, including the SGLT2 inhibitors (SGLT2is), GLP-1 receptor agonists (GLP-1 RAs) and the mineralocorticoid receptor antagonist (MRA) finerenone decrease the risk for major adverse cardiovascular events, hospitalization for heart failure and adverse kidney outcomes in patients with T2D at high risk for CVD and CKD. The mechanisms for these beneficial effects are not known, but they appear independent of glucose lowering. Although these trials excluded patients with known T1D, between 45-60% of patients were treated with insulin. Applying earlier estimates of IAA prevalence in T2D, it is likely that up 10% of patients in these trials were IAA positive and probably had T1D. Several of these trials collected plasma and serum samples in biorepositories. Thus, it is possible to retrospectively identify IAA positive patients in these trials to determine, in subgroup analyses, whether there are any safety signals or early signals of efficacy of these drugs in patients with autoimmune diabetes, before embarking on large, expensive, dedicated trials of these drugs in patients with T1D.
We propose 2 specific aims: 1) To measure, in a staged approach, IAA (GAD, IA2, ZNT8) and c-peptide in ~9400 biorepository samples from 4 T2D CVOTs that have tested SGLT2is, GLP-1RAs or finerenone. This will tell us how common autoimmune diabetes is in these high-risk populations as well as whether there is any indication of a difference in safety or efficacy for these drugs in those with probable T1D, relative to the trial population as a whole. 2) We will then use data from these studies, as well as other bioinformatic approaches, to classify patients in all CVOTs with publicly available data as to their likelihood having autoimmune diabetes and perform similar subgroup analyses in this broader patient population to increase the strength of any safety and efficacy signals. This approach will allow us to provide preliminary data on the safety and efficacy of these three classes of drugs in patients with autoimmune diabetes in a rapid and cost-efficient manner. While these analyses will not obviate the need for future trials, findings of similar safety and efficacy profiles in IAA positive patients will support further development of SGLT2i, GLP-1RA and/or finerenone for CVD and CKD risk reduction in patients with T1D and allow prioritization and focus on the most promising candidates.
Anticipated Outcome
Hypothesis: We hypothesize that measurement of islet autoantibodies (IAA) and c-peptide in biorepository samples from patients enrolled in T2D CVOTs will identify 5-10% of patients with adult-onset autoimmune diabetes. In subgroup analyses, patients with probable T1D in these and other trials will have a similar safety and efficacy profile compared to the general trial population, as well as propensity-matched controls.
To test this hypothesis, we will pursue 2 Specific Aims:
1) To measure IAA (GAD, IA2, ZNT8) and c-peptide in ~9400 patients from 4 T2D CVOTs representing 3 classes of medications: SGLT2is (sotagliflozin, ertugliflozin), GLP-1RAs (semaglutide) and MRAs (finerenone). These data will be used to determine the prevalence of IAA positivity among high-risk patients in these CVOTs and to perform subgroup analyses comparing the safety and efficacy profile among patients with probable T1D to those with typical T2D.
2) Using data from these analyses as well as other bioinformatic approaches, to classify patients in all CVOTs with publicly available data as to their likelihood having probable T1D and perform similar subgroup analyses in a broader patient population to increase the strength of any safety and efficacy signals.
Outcomes: We expect that the prevalence of any IAA positive among patients in these trials will be at least 5% and likely closer to 10%. The primary analyses will compare the safety and efficacy of the treatments in each trial in subgroups defined by IAA status (any + vs -). Multiple sensitivity analyses will be performed including subgroups defined by multiple IAA (+ vs -), IAA “titer” (high vs low vs -), IAA + with low C-peptide < 0.1 mg/mL) and a clinical diagnosis of “probable T1D” based on IQVIA or similar algorithms in combination with IAA status. The latter analyses will address concerns about the significance of IAA vs “true” adult onset T1D, which would be more likely with more stringent criteria. Safety outcomes will include the incidence and total rates of severe hypoglycemia events, DKA and other adverse events of interest (nausea, vomiting, genital mycotic infections, urinary tract infections, hypovolemia, hyperkalemia). Efficacy outcomes will include time to first MACE and the individual components of MACE, time to first HHF, time to first composite kidney end point (a sustained 40% decline in eGFR with/without albuminuria, kidney death, CV death) as well as HbA1c, body weight and the annualized decrease in eGFR and urine albumin/creatine ratio.
Significance: Findings of similar safety and efficacy profiles in IAA positive patients will support further development of SGLT2i, GLP-1RA and/or finerenone for CVD and CKD risk reduction in patients with T1D.
Relevance to T1D
Relevance to Persons with T1D: At present, despite the high risk of cardiovascular disease (CVD) and chronic kidney disease (CKD) in patients with T1D, there have been few dedicated studies testing interventions to reduce risk for these adverse outcomes in this population. In part, this is because of the perceived difficulty in enrolling these studies, the high costs of these long-term trials, the lack of an FDA mandate (as was the case for T2D) and the relatively small market. Consequently, apart from glycemic control, the approach to CVD and CKD risk reduction in T1D has largely been extrapolated from studies in individuals with T2D.
We anticipate that 5-10% of persons in T2D cardiovascular outcomes trials (CVOTs) will be islet autoantibody (IAA) positive, indicating that they probably have T1D and not T2D. We also anticipate that the safety and efficacy profile in persons with probable T1D in these trials will be similar to that in patients with typical T2D. If the safety and efficacy profiles are, indeed, similar in IAA positive patients compared to those with T2D, clinicians may be more willing to prescribe these medications off-label to reduce risk for CVD and CKD in patients with T1D. Such findings would also support further development and dedicated trials of SGLT2i, GLP-1RA and/or finerenone for CVD and CKD risk reduction in patients with T1D. Conversely, if these analyses identify unique risks of these drugs in patients with probable T1D, clinicians would likely be dissuaded from using these drugs in patients with T1D, despite the apparent benefits in patients with T2D. In either case, results from this study will allow us to provide preliminary data on the safety and efficacy of these three classes of drugs in patients with autoimmune diabetes in a rapid and cost-efficient manner. While these analyses will not obviate the need for future trials, they may allow prioritization and focus on the most promising candidates.