Atypical Antipsychotic Drugs and Diabetes Mellitus in the US Food and Drug Administration Adverse Event Database: A Systematic Bayesian Signal Detection Analysis

Ross A. Baker; Andrei Pikalov; Quynh-Van Tran; Tatyana Kremenets; Ramin B. Arani; P. Murali Doraiswamy

Disclosures

Psychopharmacol Bull. 2009;42(1):1-21. 

In This Article

Results

Overall Reports of Diabetes-related Adverse Events (DRAEs)

There were a total of 8,032 cases of the 24 MedDRA-defined DRAEs for individuals receiving atypical antipsychotics or haloperidol. Among the study drugs, olanzapine and clozapine demonstrated the highest frequencies of DRAEs (3,620 and 1,299 cases, respectively), followed by risperidone (1,163 cases), quetiapine (869 cases), haloperidol (546 cases), aripiprazole (295 cases) and ziprasidone (240 cases).

Diabetes mellitus (2,784 cases) was the most frequently reported DRAE, followed by hyperglycemia (1,347 cases), increased blood glucose (1,104 cases), diabetic ketoacidosis (782 cases), and non-insulin-dependent diabetes mellitus (707 cases). Insulin-dependent diabetes mellitus was reported in 234 cases, diabetic coma in 269 cases and glycosuria in 90 cases. Table 2A shows the number of DRAEs by each of the antipsychotics for the six most clinically relevant adverse events. The incidence of these six events per drug is similar to the incidence across all 24 DRAES; olanzapine and clozapine demonstrated the highest frequencies, followed by risperidone, quetiapine, haloperidol, aripiprazole and ziprasidone. Table 2B shows the prescription volume for each antipsychotic from the time of launch to the end of 2006. It also shows the overall number of AEs reported for each drug in the AERS database.

Adjusted RRs for DRAEs

In total, 14 DRAEs had elevated EB05 values (≥2), indicating a 95% likelihood of DRAEs occurring at least two times more frequently than expected, including: diabetes mellitus (including insulin dependent or non-insulin dependent); diabetic coma (including hyperglycemic or hyperosmolar coma); diabetic ketoacidosis; gestational diabetes; glucose tolerance impaired; glucose urine present; glycosuria; hyperglycemia (including hyperosmolar nonketotic syndrome) and insulin resistance ( Table 2A , Table 2B , Table 2C and supplementary Table 3 ). All seven study drugs demonstrated an EB05 score ≥2 for at least one DRAE. Overall, olanzapine was associated with the most DRAEs, with 13/24 DRAEs having elevated EB05 values. Furthermore, EB05 values were highest for drug-event combinations with olanzapine and five DRAEs had EB05 values ≥9 (non-insulin dependent diabetes mellitus; diabetic coma; diabetic ketoacidosis; glycosuria; and hyperglycemic hyperosmolar nonketotic syndrome).

In Figure 1a, adjusted RRs (EBGM values and 90% CIs) are plotted against logistic regression ORs for diabetes mellitus, which is the most frequently reported DRAE. Logistic regression odds ratios (and 90% CIs) for diabetes mellitus were: olanzapine 16.0 (15.3-16.8); clozapine 4.0 (3.7-4.3); risperidone 2.9 (2.6-3.2); quetiapine 2.9 (2.4-3.3); ziprasidone 1.7 (1.2-2.2); aripiprazole 1.3 (0.9-1.7) and haloperidol 1.3 (1.0–1.5). Based on both analytic methods, olanzapine separates with the highest risk; quetiapine, risperidone, and clozapine all demonstrate moderate risk; whereas ziprasidone, haloperidol, and aripiprazole demonstrate the least risk by falling at or below the threshold of 2 for both EB05 and LR05. The cumulative number of diabetes mellitus reports (Figure 1b) indicate that EB05 values of two or above for diabetes mellitus were evident from 1998 for clozapine and 1999 for olanzapine (the second year of its introduction in the AERS database).

Figure 1A.

Adjusted Reporting Ratio (EBGM Score) and 90% Confidence Intervals (EB05–EB95) Versus Logistic Regression Odd-ratios (LROR) and 90% Confidence Intervals (LR05–LR95) for Diabetes Mellitus

Figure 1B.

Cumulative Number of Drug–event Reports for Diabetesmellitus for Atypical Antipsychotic Drugs and Haloperidol

As observed with diabetes mellitus, olanzapine showed the highest EBGM value for hyperglycemia (supplementary Table 3 ). Whereas hyperglycemia was the second most reported DRAE, only olanzapine had an EB05 value >2, suggesting a potential signal. Haloperidol showed the lowest EBGM value. Blood glucose increase was the third most common DRAE; however, none of the atypical antipsychotics or haloperidol showed elevated EBGM values with EB05 >2, suggesting no association between agents and this DRAE (data not shown).

Although the cumulative frequencies for diabetic ketoacidosis and non-insulin-dependent diabetes mellitus were generally lower than for other DRAEs, EBGM EB05 values for these DRAEs exceeded the threshold of a potential signal (>2) for several agents ( Table 2C ). Haloperidol, risperidone, quetiapine, clozapine and olanzapine all exceeded the threshold for diabetic ketoacidosis; ziprasidone and arip-iprazole did not. All agents had elevated EBGM values for non-insulin-dependent diabetes mellitus.

The cumulative number of reports of diabetic ketoacidosis (Figure 2) indicate that olanzapine demonstrated EB05 values >2 from the first years of inclusion within the database. By 2006, associations for olanzapine and diabetic ketoacidosis approached EB05 values of ≥8 (Figure 2). Neither aripiprazole nor ziprasidone showed elevated reports of diabetic ketoacidosis.

Figure 2.

Cumulative Number of Drug–event Reports for Diabetic Ketoacidosis for Atypical Antipsychotic Drugs and Haloperidol

Drug–diabetes Associations by Age Group

Children and Adolescents (Less Than or Equal to 17 Years). A total of 258 cases of the 24 DRAEs were identified for individuals, less than 17 years of age, receiving atypical antipsychotics or haloperidol. Among the study drugs, olanzapine and risperidone demonstrated the highest frequencies of DRAEs (82 and 56 cases, respectively).

Of the 24 DRAEs identified, hyperglycemia (61 cases) was the most frequently reported DRAE in this age group, followed by diabetes mellitus (58 cases) and increased blood glucose (37 cases). The number of diabetes events and events with EB05 values (≥2) are shown in Figure 3.

Figure 3.

Association of Diabetes Mellitus With Age for Atypical Antipsychotic Drugs and Haloperidol

In this age group, three DRAEs had elevated EB05 values (≥2), indicating a 95% likelihood of DRAEs occurring at least two times more frequently than expected, including: diabetes mellitus (olanzapine and clozapine); non-insulin-dependent diabetes mellitus (olanzapine and quetiapine); and diabetic coma (olanzapine).

Adults (18-64 Years). A total of 5,764 cases of the 24 DRAEs were identified for individuals, aged 18-64 years, receiving atypical antipsychotics or haloperidol. Results for this group were similar to the total population; olanzapine and clozapine demonstrated the highest frequencies of DRAEs (2,500 and 1,115 cases, respectively) and all seven study drugs demonstrated an EB05 score ≥2 for at least one DRAE.

Of the DRAEs, diabetes (1,825 cases) and hyperglycemia (955 cases) were the most frequently reported DRAEs in this age group. Olanzapine, quetiapine, risperidone, clozapine and ziprasidone all demonstrated elevated EBGM values for diabetes mellitus (Figure 3), whereas only olanzapine demonstrated 2 ≤ EB05 < 4 for hyperglycemia in this age group (Figure 4).

Figure 4.

Association of Hyperglycemia With Age for Atypical Antipsychotic Drugs and Haloperidol

Elderly (Over 65 Years). A total of 529 cases of the 24 DRAEs were identified for individuals, over 65 years of age, receiving atypical antipsychotics or haloperidol. Among the study drugs, olanzapine and risperidone demonstrated the highest frequencies of DRAEs (243 and 99 cases, respectively).

Of the 24 DRAEs identified, diabetes (176 cases) was the most frequently reported DRAE in this age group, followed by hyperglycemia (122 cases) and increased blood glucose (116 cases). The number of events and events with EB05 values (≥2) for diabetes are shown in Figure 3.

In this age group, three DRAEs had elevated EB05 values (≥2), indicating a 95% likelihood of DRAEs occurring at least two times more frequently than expected, including: diabetes mellitus; non-insulin-dependent diabetes mellitus; and diabetic coma (all with olanzapine only).

Diabetes Mellitus Associations in AERS

Among more than 4,000 drugs in the entire AERS database (analyses using generic drug names, ATC Level 5), olanzapine had the second highest frequency of diabetes reports (1,306 cases) and the highest EBGM adjusted RR (9.59) and 90% CIs (9.16-10.03). Clozapine was the fifth highest.

processing....