Latent autoimmune diabetes in the adults (LADA) in Asia: from pathogenesis and epidemiology to therapy
Professor Paolo Pozzilli
Department of Endocrinology and Diabetes
University Campus Bio Medico
Via Álvaro del Portillo, 21
00128 Rome, Italy.
Diabetes mellitus is a metabolic disorder resulting from a defect in insulin secretion, insulin action or both. An effect of this process is chronic hyperglycaemia with disorder of carbohydrate, fat and protein metabolism and with long-term complications of diabetes including retinopathy, nephropathy and neuropathy. Latent autoimmune diabetes in adults (LADA) is a type of autoimmune diabetes that resembles Type 1 diabetes (T1D), however, it shows a later onset and slower progression towards insulin necessity. Epidemiological studies suggest that LADA may account for 2–12% of all cases of diabetes in adult population. The epidemiology and phenotypic characteristics of LADA may vary between Caucasian and Asian diabetic patients as lifestyle, food habits and body mass index differ between these two populations. Data on LADA from population-based studies in Asia are sparse and only few studies have looked at it.
A number of attractive therapeutic interventions may be envisaged for prevention of beta-cell loss in LADA, including hypoglycaemic and immunomodulatory agents. Because the autoimmune process in LADA seems to be slower than in childhood T1D, there is a wider window of opportunities for intervention. In deciding the best therapeutic approach, features of LADA should guide therapy including presence of other comorbidities that may influence the therapeutic choice.
Introduction on pathogenesis of LADA
Autoimmune diabetes is characterized by the presence of one or more islet-specific autoantibodies, including islet cell autoantibodies (ICA) and autoantibodies directed against the three major islet autoantigens, including glutamic acid decarboxylase (GAD), protein tyrosine phosphatase IA-2 (IA-2A) and its isoform IA-2 (IA-2) and insulin autoantibodies (IAA) [1–4]. Type 1 diabetes (T1D) may occur at any age, in young individuals before and soon after adolescence, during middle age life or even in the elderly. It is uncommon to diagnose T1D in the adult age and when this occurs, the detection of one autoantibody helps to identify a subgroup of patients with type 2 diabetes (T2D) with clinical characteristics more similar to those with classical T1D predictive of a rapid progression to insulin requirement . Among patients diagnosed with T2D some of them develop insulin-requiring diabetes and some of these patients can be identified earlier in the natural history of the disease by the presence of circulating islet autoantibodies .
Latent autoimmune diabetes in adults is a form of autoimmune diabetes that resembles T1D, however, it shows a later onset and slower progression towards insulin requirement [7–9]. Alternative terms that have been used to describe this condition include type 1.5 diabetes, latent T1D, slowly progressive Insulin Dependent Diabetes Mellitus (SPIDDM)  or youth onset diabetes of maturity. All these terms are often considered synonymous based on the fact that this form of diabetes is a type of disease in adults with an autoimmune basis eventually leading to insulin therapy for its treatment.
When diagnosed in adults this form of diabetes is characterized by the presence of GAD and IA2 autoantibodies (the latter less common), very rarely insulin antibodies (IA) and by the occurrence of a less high-risk human leukocyte antigen (HLA) genetic susceptibility typical of T1D. Patients are usually diagnosed after 35 years of age and are often misdiagnosed as T2D. Glycaemic control is initially achieved using oral hypoglycaemic agents but soon or later patients become insulin dependent . Unfortunately, there is still an open debate regarding the definition of LADA because it is not clear if islet cell autoantibody positivity marks a distinct condition or simply a risk factor to insulin progression in a T1D phenotype. Therefore, indications regarding treatment strategies are even more difficult .
Based on GAD autoantibodies titre, LADA has also been subclassified into two forms. Patients with high GAD autoantibody levels were typed as LADA 1 with phenotypic similarities to T1D (lower C-peptide levels, lower BMI, more ketosis prone), whereas patients with low GAD autoantibody titre, subclassified as LADA 2, are less ketosis and dyslipidemia prone as compared with LADA 1. The frequency of obesity, hypertension, dyslipidemia and CHD appears lower in LADA 1 than in LADA 2 . It has also been reported that GAD autoantibody-positive LADA patients treated with insulin showed more frequently autoantibodies towards the COOH terminal (GADA-C), whereas GADA-M (autoantibodies towards the middle epitope) tend to occur more frequently in LADA patients who are on hypoglycaemic drugs and/or diet .
Another study has suggested that patients with LADA may share genetic features with both T1D and T2D patients . In addition, several studies reported on metabolic disorders, such as hypertension and obesity, and a family history of diabetes and their association with an higher risk of developing LADA [15,16].
To identify LADA between T1D and/or T2D, the Immunology of Diabetes Society has proposed three main criteria that are as follows: (a) adult age of onset (> 30 years of age); (b) presence of at least one circulating autoantibody (GAD, ICA, IAA or IA-2) and (c) insulin independence for the first 6 months after diagnosis . In a retrospective study by Fourlanos et al., five clinical parameters were found to occur more frequently in LADA than in T2D: (a) age of onset < 50 years; (b) acute symptoms (polyuria/polydypsia, weight loss); (c) BMI < 25 kg/m2; (d) personal history of other autoimmune diseases; (e) family history of autoimmune disease .
Epidemiology of LADA in Asia
Epidemiological studies suggest that LADA may account for 2–12% of all cases of diabetes in adult population . The prevalence of LADA in Western countries varies between 2.8% and 10% in subjects affected by T2D [20,21]. The Diabetes Outcomes Progression Trial reported that GAD autoantibody positivity is 4.2% in North America and 3.7% in Europe among individuals with T2D , whereas in China, two clinical studies have shown that the estimated prevalence of LADA in patients with T2D is around 7% . The epidemiology and phenotypic characteristics of LADA may vary between Caucasian and Asian diabetic patients as lifestyle, food habits and body mass index differ between these two populations. Moreover, in Caucasians, the prevalence of LADA differ consistently between patients living in north and south Europe ranging between 4% and 10% (http://www.actionlada.org).
Data on LADA from population-based studies in Asia are sparse and only few studies have looked at it. Recently Qi X. et al. showed the result of a study where they investigated the prevalence and correlates of LADA in an adult population in Tianjin, China. Prevalence of LADA in this group of T2D subjects was 9% .
In 2007 a study was designed to determine the prevalence of LADA in a larger UK resident South Asian population and to characterize the phenotypic features and genetic basis of the disease in this ethnic group. Autoantibodies were detected in 13 of 500 (2.6%) individuals with T2D [of whom 8 were GAD65 positive (1.6%) and 6 were IA-2 positive (1.2%), including 1 subject who was positive for both] and 8 of 206 (3.9%) control subjects [of whom 3 were GAD65 positive (1.5%) and 5 were IA-2 positive (2.4%)]. There was no significant difference in antibody titres between diabetic and control subjects . This study showed that islet cell autoimmunity was considerably less common among T2D individuals of Punjabi ancestry in Birmingham, UK, than in those of white Caucasian origin. The differences observed between the two ethnic groups may reflect both the higher prevalence of classical T2D in South Asians and their lower susceptibility to autoimmune disease.
Japan is among countries with the lowest incidence rate of childhood T1D in the world, averaging 2.4 cases/100 000/year. However, it appears that the prevalence of T1D in adulthood occurs twice compared with childhood patients . The prevalence of GAD autoantibodies in adulthood patients with T2D without insulin therapy is 3–4%, but is higher in patients with shorter duration of diabetes . Although high levels of GAD autoantibodies are one of the predictive markers for future insulin requirement, there are a certain number of patients with high GAD autoantibody titres who do not progress to insulin dependency for many years.
Lower BMI, waist/hip ratio, lower total cholesterol and triglycerides levels but higher HDL cholesterol have been reported in LADA than in patients with T2D. The prevalence of hypertension was also lower in LADA patients [10,11]. There was no significant difference between LADA and adult onset T1D in this respect.
Recently, a collaborative project between Italy and Korea has evaluated the prevalence of LADA using the same ascertainment criteria. A striking similar prevalence of LADA was reported (4.4% in Italy and 4.4% in Korea) .
A collaborative project between Italy and Korea to evaluate prevalence and features of LADA in Caucasian and Asian subjects
In a joint collaborative ongoing project between Italy and Korea, we have compared the anthropometric, clinical and metabolic data of Caucasian versus Asian LADA patients. Two large studies on the prevalence of LADA are carried out in Italy and South Korea (Non Insulin Requiring Autoimmune Diabetes [NIRAD] Study and Korean National Diabetes Program) whose overall results have been already published [29,30]. A cross-sectional analysis was performed in the 193 LADA patients of NIRAD study diagnosed in Italy and in the 39 LADA patients diagnosed in Korea. The same ascertainment criteria were applied for diagnosis of LADA in both groups of patients . Biochemical parameters were measured centrally in both countries. GAD autoantibodies were evaluated in two centralized laboratories one in Italy and one in Korea which both participated in the DASP autoantibody Workshop. Data were analysed using one-way anova test. We found a very similar prevalence of LADA in the two populations (4.4% in Italy and 4.4% in Korea). The mean age at diagnosis of LADA was nearly identical between Caucasian and Asian patients (50.3 years ± 12 versus 49.6 years ± 8.3 respectively), as well as the HbA1c values (7.5% ± 1.7 versus 7.4% ± 1.5 respectively). There was a significant higher BMI in Caucasian versus Asian LADA (27 ± 5.1 versus 25.3 ± 3.3, respectively, P < 0.01), and higher HDL (but not total) cholesterol levels (50 mg/dl ±3 mg/dl versus 44.7 mg/dl ± 11 mg/dl, respectively, P < 0.03). By contrast, triglyceride levels were significantly higher in Asians versus Caucasian (201 mg/dl ± 169 mg/dl versus 144 mg/dl ± 104 mg/dl, respectively, P < 0.01), likely reflecting differences in diet habits. This is the first study which evaluated characteristics of LADA among Caucasian and Asian diabetic patients using the same ascertainment criteria indicating that, despite a difference in prevalence of T1D, LADA prevalence is remarkably similar between these two different ethnicities. However, metabolic features differ substantially between the two ethnicities  (Table 1).
Table 1. Features of LADA in Caucasian and Asian patients (ref. 28)
|Age at diagnosis (years)||50.3 ± 12||49.6 ± 8.3||NS|
|BMI (kg/m2)||27 ± 5.1||25.3 ± 3.3||< 0.01|
|HbA1c (%)||7.5 ± 1.7||7.4 ± 1.5|| |
|Fasting blood glucose (mg/dl)||168.1 ± 58||134 ± 34||< 0.001|
|HDL Cholesterol (mg/dl)||50 ± 3||44.7 ± 11||< 0.03|
|Triglycerides (mg/dl)||144 ± 104||201 ± 169||< 0.01|
Therapy of LADA
No treatment guidelines for LADA have been published therefore patients are mostly treated as affected by T2D. However, the potential best therapeutic option for LADA patients should aim not only to obtain good metabolic control but also to allow better preservation of residual beta-cell function. In fact it has been shown that maintenance of some endogenous insulin production is associated with improved metabolic control and better long-term disease outcome . Although there is a good proportion of patients with LADA, few intervention trials have been carried in this form of diabetes  (Table 2).
Table 2. Features of study drugs used to treat Asian LADA patients
|Metformin||No evidence||No evidence||---|
|Incretin||No evidence||No evidence||---|
Diet treatment in LADA is similar to that of classical T1D. Obese LADA patients benefit from restriction in calorie consumption and increased levels of physical activity [7,32].
Sulfonylureas (SU) are one of the most common drugs used for the treatment of T2D and its mechanism of action is based on the binding to a specific site on the ATP-sensitive (Adenosine triphosphate) K+ channels at the level of plasma membrane, which leads to their closure and subsequent opening of the calcium channels and activation of an “effector” system of insulin release .
There are two randomized controlled trials (RCTs) conducted in Japan comparing glibenclamide with insulin treatment in LADA patients. The first was a pilot RCT examining insulin alone versus glibenclamide alone . Baseline HbA1c was 7.8% and 8.5% in the insulin compared with SU groups respectively. Results at 30 months of follow-up showed a significant difference in the two groups with an end of study HbA1c of 7.3% and 11.2% respectively. Finally, the SU group showed worsening metabolic control and a progressive deterioration of beta-cell function (during the study period serum-stimulated C-peptide ratio decreased almost 40% from baseline).
The second one (the Tokyo study)  was a multicenter, randomized, nonblinded clinical study that suggests that glibenclamide accelerates (or at least does not protect against) progressive beta-cell failure and is similar to (or worse than) insulin in obtaining good metabolic control (see chapter below). Therefore, SU should not be used as first-line therapy in patients with LADA
One of the three criteria to define LADA is the absence of insulin requirement for at least 6 months after diagnosis. Therefore, early insulin treatment could be a contradiction, but the rationale of this approach would be to improve metabolic control while protecting beta-cell function. The exact mechanism for the apparent beneficial effects of insulin treatment reported in several studies is not yet fully understood. Uppermost, it is thought that administration of exogenous insulin would allow beta-cell rest, at least in part by downregulating the beta-cell metabolism and/or by releasing them from the hyperglycaemic stress . Result of this is a decrease in the severity of “insulitis” and in the number of infiltrative antigen-presenting cells in and around the pancreatic islets . Another potential explanation would be that exposure to exogenous insulin would actually promote Th2 immunity in humans by an increase in IgG1 and IgG4-IA (autoantibodies to insulin) and induce an activation of insulin-specific regulatory T-cells (Tregs) [38,39]. Finally, as insulin is a major autoantigen in T1D (mainly in type 1A), immunization with exogenous insulin could determine immune modulation possibly by tolerance induction or “bystander” suppression of autoreactive T-cells through the local release of regulatory cytokines . Such evidence is based on studies in the NOD mouse, the BB rat and in a human pilot trial that demonstrated that parenteral insulin therapy protects against T1D [41,42].
In the Japanese trial, the 2-h blood glucose level during the 100-g oral glucose tolerance test tended to decrease from the baseline values, the insulin-treated group had an increased stimulated C-peptide at 30 months, but the HbA1c remained unchanged 30 months later .
To test the hypothesis that insulin therapy rather than sulfonylurea treatment is preferable to reverse or preserve beta-cell function among patients with slowly progressive insulin-dependent (type 1) diabetes SPIDDM or latent autoimmune diabetes in adults, Maruyama et al. planned a multicenter, randomized, nonblinded clinical study (the Tokyo study) that screened 4089 noninsulin-dependent diabetic patients for GAD autoantibodies . Sixty GAD autoantibody-positive noninsulin-requiring diabetic patients with a 5-year or shorter duration of diabetes were assigned to either the SU group (n = 30) or the insulin group (n = 30). Serum C-peptide responses to an oral glucose tolerance test performed yearly was followed up for a mean of 57 months. The primary endpoint of the study was an insulin-dependent state defined by the sum of serum C-peptide values during the oral glucose tolerance test (SigmaC-peptide) to be < 4 ng/ml (1.32 nM). The authors demonstrated that the progression rate to an insulin-dependent state in the insulin group was lower than that in the SU group. Longitudinal analysis showed that the sum of serum C-peptide values during the oral glucose tolerance test was better preserved in the insulin group than in the sulfonylurea group.
Other two studies carried out in Japan demonstrated preservation of beta-cell function with insulin compared with SU in ICA positive and GAD autoantibody-positive phenotypic T2D subjects [43,44].
Insulin sensitizers (metformin, thiazolidinediones)
Patients with LADA showing insulin resistance (metabolic syndrome) could benefit from therapy with metformin, a drug that improves peripheral insulin action and indirectly protects beta cells from continuous hyper stimulation. Other drugs that belong to the insulin sensitizers family are thiazolidinediones (TZDs). These compounds are very interesting for LADA because they improve insulin content and secretion, preserve beta-cell mass and islet structure, have anti-inflammatory effects, protect beta cells from oxidative stress and apoptosis and even facilitate beta-cell proliferation [45–49].
Unfortunately no studies evaluated metformin for treatment of LADA .
A randomized control trial comparing rosiglitazone plus insulin versus insulin alone in LADA patients was carried out over a total follow-up period of 18 months . LADA patients, with a fasting C-peptide of 0.3 nM or more, were enrolled and randomly assigned to receive subcutaneous insulin alone or rosiglitazone plus insulin to compare the impact on islet beta-cell function. Islet beta-cell function was evaluated by C-peptide after 2 h 75-g glucose load (PCP) and Delta CP (Delta C-Peptide = PCP-Fasting C-Peptide). During 6 months follow-up, there were no significant changes for DeltaCP and PCP levels in both groups. PCP and DeltaCP levels in insulin + RSG (rosiglitazone) group patients stayed steady during the 12 months’ observation, whereas in the insulin alone group, both FCP and PCP levels decreased significantly. PCP and Delta CP differences between 12th month and baseline were higher in insulin + RSG group than those in the insulin group. When observed up to 18 months, PCP and DeltaCP levels in insulin + RSG group patients still stayed steady, whereas PCP and DeltaCP levels decreased more in the insulin alone group.
This pilot trial seems to demonstrate that rosiglitazone combined with insulin may preserve islet beta-cell function in LADA patients even if rosiglitazone plus insulin did not improve metabolic control significantly more than insulin alone.
Incretins are a group of gastrointestinal hormones that cause glucose-dependent enhancement of insulin secretion. They also slow the rate of absorption of nutrients into the blood stream by reducing gastric emptying and may directly reduce food intake . They also inhibit glucagon release from islet alpha cells and have tropic effects on pancreas as they modify the susceptibility to apoptotic injury and stimulate beta-cell proliferation/islet neogenesis from precursor cells [49,53–56].
In T1D (and also in LADA) autoimmune T-cells attack on insulin-producing beta cells eventually causing hyperglycaemia. The common view is that to prevent and treat this disease, one must interfere with the autoimmune process. To avoid hazardous side effects, the intervention should not cause a generalized immune suppression but favour an immune modulation. The autoimmune process can be mitigated inducing tolerance to autoantigens, targeting them through the administration of autoantigen and deviation of the Th1 phenotype of antigen-reactive cells towards a Th2 phenotype. Antigens that have been used so far as tolerogens in LADA are insulin, GAD, heat shock protein (HSP) and their constituent peptides. Critical issues for a successful outcome include variables such as HLA, age at diagnosis, metabolic control and the residual beta-cell function present at diagnosis .
HSP60 is a ubiquitous protein, part of a highly conserved family of intracellular chaperones, also located in the mitochondria and mature insulin secretory granules of pancreatic beta cells, with relevant regulatory role in the innate immune system  and considered as an important autoantigen in T1D. The dominant epitope of HSP60 was found to be the HSP277 peptide, and its modified form, Diapep277 (generated to increase its stability in vivo), has been used in patients with recent-onset T1D for prevention of further beta-cell loss [59,60]. It activates anti-inflammatory effector cells through TLR2 leading to a shift from an inflammatory to a regulatory immune response.
In 2001, Raz et al. showed that patients with recent-onset T1D and basal C-peptide concentrations above 0.1 nM treated with subcutaneous injections of 1 mg p277 and 40 mg mannitol in vegetable oil (DiaPep277) tend to preserve endogenous insulin production (at 10 months, mean C-peptide concentrations had fallen in the placebo group but were maintained in the DiaPep277 group, and need for exogenous insulin was higher in the placebo than in the DiaPep277 group). In the same trial, T-cell reactivity to HSP60 and p277 in the DiaPep277 group was associated with an enhanced T-helper-2 cytokine phenotype .
A phase II double-blind multicenter RCT was conducted in 60 patients with LADA, 30–50 years old and within 2–60 months after diagnosis for evaluation of safety, tolerability and clinical, metabolic and immunological efficacy of multiple subcutaneous doses of DiaPep277. Results have not been published, but a brief report suggests good safety and tolerability, and lymphocyte response to DiaPep277 in treated patients with generation of a Th2 cytokine phenotype .
The 64-kDa pancreatic beta-cell autoantigen, which is a target of autoantibodies associated with early as well as progressive stages of beta-cell destruction, was identified as the gamma-aminobutyric acid-synthesizing enzyme glutamic acid decarboxylase. GAD65 is mainly found in beta cells and other tissues, and it is considered a major autoantigen in autoimmune diabetes. GAD65 autoantibodies are found in 70–75% of T1D patients and are considered the most sensitive autoantibody marker in LADA . Treatment with exogenous GAD induces tolerance to GAD65 by modulating the immune response in a discrete antigen-specific fashion (vaccination type of approach).
In 2005, a phase II randomized, double-blind, placebo-controlled, dose-escalation clinical trial was carried out in a total of 47 LADA patients to evaluate if alum-formulated human recombinant GAD65 (Diamyd) was safe and may influence the decline over time of beta-cell function . Fasting C-peptide levels at 24 weeks were increased compared with placebo in the 20 mcg arm group. In addition, both fasting and stimulated C-peptide levels increased from baseline to 24 weeks in the 20 mcg dose group. These changes were accompanied by an increase of the purported Tregs subsets (CD4-CD25/CD4-CD25 cell ratio) in the peripheral blood. No change in HbA1c or plasma glucose or decrease in beta-cell function was observed in any of the dose groups and no study-related adverse effects were reported.
A study in paediatric T1D subjects indicated that GAD-alum treatment had no significant effect on FCP after 15 months. After 30 months, FCP and stimulated C-peptide showed a significantly smaller decline compared with placebo .
In a recent study, Wherrett DK et al.  assess whether immunization with GAD formulated with aluminium hydroxide (GAD-alum) would preserve insulin production in recent-onset T1D. The primary outcome was the baseline-adjusted geometric mean area under the curve of serum C-peptide during the first 2 h of a 4-h mixed meal tolerance test at 1 year. Secondary outcomes included changes in HbA1C, insulin dose and safety. The authors showed that antigen-based immunotherapy therapy with two or three doses of subcutaneous GAD-alum across 4–12 weeks did not alter the course of loss of insulin secretion during 1 year in patients with recently diagnosed T1D. So it appears that GAD therapy has little effect in T1D, but may have a beneficial effect in LADA, however, more studies are needed in this group of patients.
LADA includes a significant proportion of all patients affected by diabetes. A number of attractive therapeutic interventions may be envisaged for prevention of beta-cell loss in LADA including hypoglycaemic and immunomodulatory agents. Because the autoimmune process in LADA seems to be slower than in childhood T1D, there is a wider window of opportunities for intervention. In deciding the best therapeutic approach, features of LADA (considering the many clinical variables) should guide therapy including presence of other comorbidities that may influence the therapeutic choice.
Work on LADA in our centre is supported by the Ministry of Health (Ricerca Finalizzata) and by Centro Internazionale Studi Diabete.
Conflicts of interest
None to declare.