Pharmacogenetic Profiling of Antipsychotics-Induced X-Syndrome and Diabetes (PAXD)

Specific Aims Although our original interests were condensed in atypicals, the fact that some hospitals discourage the use of atypical antipsychotics for new onset schizophrenia for policy reasons directs us to a study design looking at both types of therapeutics. Coincidently, such a design may provide some hints to the unresolved research issues mentioned above. The proposed study will bridge these important information gaps, and provide data contributing towards a better understanding of the risk and management of diabetes associated with the use of both types of antipsychotics. As some researchers in this field have described the diabetic related conditions in a broader term called "metabolic X-syndrome", a risky condition leading to cardiovascular diseases and diabetes, with insulin resistance as the major outcome, associated with two of the following conditions: truncal obesity (deposited in the thorax and abdomen, instead of the hips and thighs), high triglycerides, high LDL cholesterol or hypertension,3 we therefore extend our study to both diabetes and metabolic X-syndrome as a phenotype for further genetic study. This proposal aims to study: 1. the prevalence of diabetes/metabolic X-syndrome associated with the use of antipsychotics in Taiwanese Han Chinese populations; 2. clinical and sociodemographic correlation of diabetes/metabolic X-syndrome associated with the use of antipsychotics; and 3. the putative role of genes in the development of treatment emergent diabetes/metabolic X-syndrome secondary to the administration of antipsychotics. Background Affecting approximately 1% of any population worldwide, schizophrenia represents a major cause for morbidity and dysfunction for the afflicted and their loved ones, as well as a significant burden for the health care system and the society in general. Although antipsychotics are often remarkably effective in ameliorating schizophrenic symptoms and in enhancing social functioning, their use often is limited by their propensity to cause serious side effects that could further compromise patients' health conditions. One such side effect is diabetes. Research publications from 1919, long before the advent of psychotropic medication, had showed schizophrenia might predispose individuals to diabetes.1 Following the introduction of chlorpromazine, the first antipsychotic in the modern era, a large number of case reports of antipsychotics-induced diabetes were reported. This complication seems more serious in the new type of antipsychotic, which is generally regarded as an atypical antipsychotic due to the lack of extrapyramidal reactions seen in schizophrenics treated with antipsychotics discovered previously. Several retrospective studies based on health services claim databases, strongly suggested that psychiatric patients exposed to atypical antipsychotics may be at a higher risk for developing diabetes. In addition, patients treated with these medications also appear to be at a higher risk for developing ketoacidosis. Risk factors included ethnicity (African Americans, Asians, Hispanics and Native Americans), family history of diabetes and overweight or obesity. The association between these atypical antipsychotics and the onset of diabetes is further strengthened by observations of: 1. the time sequence between the initiation of antipsychotic treatment and the onset of diabetes; 2. remission after the discontinuation of medications; and 3. re-emergence of diabetes following the re-introduction of atypical antipsychotics. Together, these data indicate that treatment emergent diabetes, along with other metabolic disturbances (e.g., dyslipidemia and obesity), represents a serious issue in the use of atypical antipsychotics. This notwithstanding, many questions of crucial clinical and theoretical interest remain un-clarified. Major current debates and unresolved research issues include: (1) schizophrenia per se, versus the use of antipsychotics, in triggering diabetes; (2) whether there are differences between "typicals" and "atypicals" in such an effect; (3) whether there are differences among different "atypicals;" (4) whether, and to what extent, treatment emergent diabetes may be associated with, or independent of, weight gain, which also often is associated with the use of antipsychotics; and (5) genetic and environmental risks in association with treatment emergent diabetes. This application aims at addressing some of these important issues, as well as establishing a platform for future studies that may be more comprehensive and in-depth in nature. In contrast to the plethora of case reports and reports based on secondary data analyses, there is a dearth of information deriving from studies that might be more useful for providing definitive answers to some of these questions. Our review of the extant literature led to two studies that are worth noting in this regard: 1. A five-year longitudinal study conducted by Henderson et al in Boston, reporting that patients treated with clozapine experienced significant weight gain and lipid abnormality and appear to be at an increased risk for developing diabetes.2 Overall, 36.6% (30/82) of the patients developed newly diagnosed diabetes during the 5-year period. 2. In a sizeable cross-sectional study conducted in northern Sweden, prevalence of diabetes in patients receiving clozapine was approximately 12% versus 6% in controls.3 It is of note that both studies focused only on clozapine, and neither of them provide information specifically on populations of Oriental or Asian backgrounds. Obesity and overweight are known risk factors for insulin resistance, which could eventually lead to the development of type 2 diabetes. Since most atypicals (as well as schizophrenia itself) are likely to induce significant weight gain, it is reasonable to speculate that diabetes associated with atypicals might be secondary to weight changes. However, a number of reasons suggest that factors in addition to/or other than weight gain may also be important: (1) diabetes also occurs in patients who do not gain weight;2 (2) in many cases, improvement in diabetes occurred within 2 weeks, too fast for this to be an effect of a reversal in weight changes; and (3) the relatively high rate of ketoacidosis, as well as the rapidity of the recovery following the discontinuation of the treatment, suggests that the putative diabetogenic effects of the atypicals might be secondary to the inhibition of insulin secretion, rather than insulin resistance. Very little is known about the mechanisms of typical antipsychotics in obesity and overweight, however, there should be some shared pathways in both types of antipsychotics, and specific pathways used by only typicals or atypicals. The possible role of insulin secretion in mediating atypicals' diabetogenic effects is further supported by the abundance of serotonin receptors in pancreatic beta cells. In addition to the mechanisms involving the production and function of insulin, factors affecting the pharmacokinetics and pharmacodynamics of atypicals might also significantly determine whether individuals and/or populations would be more or less likely to develop diabetes, as well as the severity of such an unintended effect. Putting these considerations together, we propose a multi-factorial working model which, while emphasizing the central role of the function of the pancreatic beta cells, also will include factors mediating insulin resistance, as well as those controlling the metabolism of the atypicals, and their effects on the central nervous system. Such a model concurrently considers genetic and other factors that might determine the effects of the atypicals and factors altering glucose homeostasis in target tissues, as well as the interactions among these factors. Previous achievements Dr. Loh has been engaged in studies of psychiatric genetics since 1993. These studies mainly emphasized the genetic susceptibility of alcoholism and illicit drug abuse. The research extended from thesis work to an international collaboration and is still being conducted using both family and case control designs. Methods Subject Recruitments: Research subjects will be recruited from three collaborative sites: Taipei City Hospital Songde Branch, Mackay Memorial Hospital and Yu-Li Hospital. The Institutional Review Boards of the National Health Research Institutes, the Yu-Li Hospital and the Mackay Memorial Hospital have approved this study, while a permit from the Institutional Review Board of the Taipei City Hospital Songde Branch is still being applied. Inpatients and outpatients diagnosed with schizophrenia according to DSM-IV criteria, treated either with typical or atypical antipsychotics, either male or female, age between 18 to 65 years old, are eligible for this study. Exclusion criteria include: 1. Illegal drug use/abuse/dependence; 2. alcoholism; 3. pregnancy; and 4. clinical diagnosis of type I or type II diabetes before the onset of schizophrenia or treatment of schizophrenia (via chart review). According to empirical experience, around 150 new onset schizophrenics may be collected each year from our three collaborative sites. For chronic schizophrenics, clinicians must be aware that the eligible patients are those who can be suitably grouped in statistics (e.g., poor responder to typical psychotics). Eligible subjects will be invited to participate in this study, after the signing of written informed consent. Possible Impediments: The patients may complain of too many biochemical measurements conducted in the study period. We can reduce the total number of biochemical measurements to 3\~4 times a year for all research subjects or for those who complain. We will lose some information but the data should be able to reveal the trend of biochemical changes. Clinical and Laboratory Procedures * Clinical Assessments for New cases Clinical information of subjects who come for the treatment of schizophrenia for the first time will be collected by using the following assessment tools in the Chinese version: (I) Clinical Global Severity (Clinical Global Impressions - Severity) or Positive and Negative Symptom Scale (PANNS): CGI Severity or PANNS is used by the clinician to record the severity of schizophrenic illness at the time of assessment. The score ranges for CGI are from 1 (normal, not at all ill) to 7 (among the most extremely ill patients). One of these scales will be chosen to access the psychiatric conditions of the patients. (II) Diabetes Risk Assessment: The Diabetes Risk Assessment (ADA, 2000) is a 7-item self-report scale designed to determine the presence of known risk factors for the development of Type 2 diabetes. Each item is assigned a point value and points are then totaled to achieve an overall risk rating. (III) Life Style Survey: The Life Style Survey is a self-report scale specifically designed for this study to ascertain certain aspects of daily health habits of the study participants. Behaviors related to smoking, alcohol use, substance use, diet, exercise, and family history are evaluated. The CGI, ADA and assessment tools in the Chinese version were translated and validated by Dr. Tsuo-Hung Lan, and have been used in his clinical studies of metabolic X-syndrome (YLH93001 and B20030902). ---Data collection for Old cases: While new cases provide the information we are interested in, patients who have been treated for schizophrenia previously and with well documented records warrant a retrospective study. Although the psychiatric and physical conditions may not consist of all the variables we are interested in, they nevertheless point out a direction for the study of new cases. In brief, psychiatric and physical information of old cases are assessed by chart review and valued variables (e.g., Brief Psychiatric Rating Scale \[BPRS\], triglyceride levels, glucose levels, etc.) and DNA will be collected and used in further pharmacogenetic studies. If possible, the variables of necessity described below will be collected by following a simplified guideline derived from the current protocol. * Physical Assessments: Weight, height, hip and waist circumference and blood pressure, will be collected by standard methods. ---Biochemistry: The following biochemical values are measured in a fasting state every two months, for 6 times: glucose, insulin, leptin, lipid panel (Total cholesterol, HDL cholesterol, LDL cholesterol and triglycerides) and glycohemoglobin (HBA1C or A1C). For exploratory purposes, the data will be processed and examined in two ways. First, we adopt diagnostic guidelines suggested by The Expert Committee On The Diagnosis And Classification Of Diabetes Mellitus, and The American Diabetes Association,5 to define and classify subjects' glucose metabolism status. However, because these criteria change from time to time and because we are interested to see how these biochemical values may form a new index, a second option to manipulate all these variables has been planned. We use the measurement of HbA1C as an additional index for diabetes. The Expert Committee On The Diagnosis And Classification Of Diabetes Mellitus did not include HbA1C as one of the criteria of diabetes because of difficulties in standardizing reference ranges of such a measurement across laboratories. This is much less a concern in this study since all procedures will be performed in the same laboratory, and the results are used for research rather than clinical purposes. The addition of the HbA1C affords us with information on glycemic levels in a time scale of weeks, whereas plasma glucose varies greatly within any given day and from day to day. Thus, an elevated HbA1C indicates a chronic state of hyperglycemia while hyperglycemia as measured by the FPG or 2-h PG may be transitory. * Pharmacogenetics: Current suggestions include candidate genes involved in pancreatic beta cell insulin secretion (e.g., HNF-1a, HNF-4a), insulin action (e.g., calpain 10, TNFα), serotonin/dopamine functions (e.g., 5-HT2A, 2C, D2, COMT) and pharmacokinetic (e.g., CYP1A2, CYP2D6,) pathways. SNPs with a population frequency of 0.1\~0.15 will be concerned as priority. ---Data Analysis: Clinical Data: The primary outcome is the diagnosis of diabetes based on American Diabetes Association criteria. Estimates of insulin secretion (fasting glucose/insulin ratios) and insulin sensitivity as well as demographic variables and evidence of over-weight/obesity will be evaluated as risk factors for the development of diabetes. Pharmacogenetic Data: The approach to analysis will be a case-control association study: SNPs within a gene will be combined to form haplotypes to provide more informative genotypes. Association studies of haplotypes will be done first, comparing the diabetic with the non-diabetic group. Modeling of contributions of multiple genes will also be conducted. When SNPs associated with antipsychotics induced diabetes are identified, their interactive effects will be analyzed by multiple logistic regression. ---Sample Size and Power Considerations: Based on current available data of the CATIE (Clinical Antipsychotic Trials of Intervention Effectiveness) conducted in the U.S.A, a 9.5% incidence rate of diabetes in schizophrenics treated with atypical antipsychotics is predicted. In addition, Hagg reported a 12% incidence rate of confirmed type II diabetes among patients treated with clozapine in their cross sectional study.4 We used the Genetic Power Calculator6 to estimate the sample size needed. We defined a power of 0.8, type I error rate of 0.05, with simple and common acceptable parameters, i.e., (1) risk ratio of susceptible homozygous genotypes set at 3, (2) risk ratio of heterozygous genotypes set at 1.5, (3) prevalence of phenotypes of interest set at 10%. When SNPs with allele frequency of 0.15 are used in such conditions, approximately 135 cases and 270 controls are needed to reach this desired power. Expectedly, to obtain 135 subjects with atypical antipsychotics-induced diabetes, we need to examine about 1350 new schizophrenic patients treated with antipsychotics. Such an expectation implies that the pharmacogenetic study of atypical antipsychotics-induced diabetes is not feasible, however, in reality, the prevalence of hyperglycemia (one of the components of X-syndrome) among patients treated with antipsychotics may be higher than 10% in a Han Chinese population. In addition, the susceptible genotype may have a higher risk ratio. Thus, posterior power estimations may give even higher power than the conservative estimation even at a smaller sample size. Moreover, our samples will merge from samples from other medical centers. Thus, the task to examine 1350 schizophrenic patients treated with antipsychotics, is feasible. ---A Sample for Preliminary Pharmacogenetic Studies: We plan to examine a collection of DNAs in which 292 samples were from the Tao-Yuan Psychiatric Hospital and 393 samples were from the Yu-Li hospital. The samples were evaluated with the same clinical procedures and physical measurements described above. However, leptin was not measured and the HbA1C was only evaluated once. Furthermore, the rest of the biochemistries were only measured twice at the 1st month and 6th month. Nevertheless, this sample collection is worth being explored in a preliminary study. Anticipated Results and Discussions Years 1 and 2: Collect clinical information, biochemical data and DNA samples, carry out preliminary statistical analysis and examine the 292 and 393 samples from the Tao-Yuan Psychiatric Hospital and Yu-Li hospital, respectively. Years 3, 4 and 5: Form composite index from preliminary statistical analysis, continue collecting DNAs and conduct pharmacogenetic studies using the DNA samples collected in this study. A better understanding of the nature and correlates of diabetic complications is crucial in the field's ongoing efforts in improving the care and welfare of schizophrenic patients. In addition, diabetes and X-syndrome associated with antipsychotics and/or schizophrenia also may represent a unique opportunity for understanding the pathogenesis and management of diabetes in general. It also could be potentially a particularly effective model for studying gene-environment interactions that are crucial for understanding most common human diseases. Ultimately, our goal is to develop genetic tests to predict therapeutic responses. As such, from the view of a national health economy, this study may prove that personalized medicine saves money. This study provides an opportunity for the researchers to be trained to extend statistical skills in estimating the putative role of genes in therapeutic responses. In addition, an understanding of the HbA1C, an interesting but debatable laboratory diagnostic test for diabetes, will be gained. References 1. Kohen D. Diabetes mellitus and schizophrenia: historical perspective. Br.J.Psychiatry Suppl 2004; 47: S64-S66. 2. Henderson DC, Cagliero E, Gray C et al. Clozapine, diabetes mellitus, weight gain, and lipid abnormalities: A five-year naturalistic study. Am.J.Psychiatry 2000; 157: 975-981. 3. Timar, O., Sestier, F. and Levy, E. (2000). Metabolic Syndrome X: A Review. Canadian Journal of Cardiology, 16:779-789. 4. Hagg S, Joelsson L, Mjorndal T, Spigset O, Oja G, Dahlqvist R. Prevalence of diabetes and impaired glucose tolerance in patients treated with clozapine compared with patients treated with conventional depot neuroleptic medications. J.Clin.Psychiatry 1998; 59: 294-299. 5. Genuth S, Alberti KG, Bennett P et al. Follow-up report on the diagnosis of diabetes mellitus. Diabetes Care 2003; 26: 3160-3167. 6. Purcell S, Cherny SS, Sham PC. Genetic Power Calculator: design of linkage and association genetic mapping studies of complex traits. Bioinformatics. 2003; 19: 149-150.