2019 Volume 66 Issue 6 Pages 571-574
Some categories of drugs are known for causing hyperglycemia or diabetes such as steroids, antipsychotics, and immunosuppressant. However, there has been little evidence from studies about the proportion of each drug in the context of drug-induced diabetes. In this study, we used data from the Japanese Adverse Drug Event Report (JADER) database, a spontaneous reporting system database maintained at the Pharmaceuticals and Medical Devices Agency (PMDA) of Japan, reported between April 2004 and June 2017. Among 459,250 reports of adverse drug reactions in JADER database, reported instances of the adverse event of hyperglycemia or diabetes were extracted. After the exclusion of anti-diabetes drugs, the drugs frequently implicated in the development of hyperglycemia or diabetes, including prednisolone, tacrolimus, everolimus, ribavirin, quetiapine, aripiprazole, interferon alfa-2b, risperidone, atorvastatin, dexamethasone, ciclosporin, nilotinib, methylprednisolone, or nivolumab, were identified. Everolimus, a mammalian target of rapamycin (mTOR) inhibitor, was manifested as the third most frequently associated drug with hyperglycemia or diabetes (340 cases), following prednisolone (694 cases) and tacrolimus (393 cases), and the reporting odds ratio (ROR 8.56, 95% CI 7.65–9.57) of this drug was higher than that of the two aforementioned drugs (ROR 3.96, 95% CI 3.66–4.28 and ROR 3.51, 95% CI 3.17–3.89). These results suggest that there is a potent association of evelolimus with hyperglycemia in clinical practice in Japan.
DRUG-INDUCED DIABETES have been known as a result of multiple mechanisms due to a variety of drugs. The Japanese Adverse Drug Event Report (JADER) database is available at the Pharmaceuticals and Medical Devices Agency (PMDA) (https://www.pmda.go.jp/).
We reviewed the JADER database to identify the drugs that can cause hyperglycemia or diabetes in clinical practice. Mammalian target of rapamycin (mTOR) inhibitors are used as immunosuppressant or antineoplastic drugs in clinical practice [1-5]. Although basic and clinical studies suggest that mTOR inhibitors can cause hyperglycemia by inducing the combination of impaired insulin secretion and insulin resistance [6-10], the reports about drug-induced hyperglycemia in the general population in Japan is still limited. In this study, we showed that everolimus, an mTOR inhibitor, showed significant reporting odds ratio to cause hyperglycemia based on JADER database.
Institutional review board approval was not required for this study, because JADER is an unlinkable anonymized database open to the public. We created a dataset for analysis from the JADER database as described previously [11]. We extracted spontaneous adverse drug reaction (ADR) reports published between April 2004 and July 2017 that were associated with systemically administered drugs. The Medical Dictionary for Regulatory Activities/Japanese version (MedDRA/J) (version 20.0) provides the Standardized MedDRA Queries (SMQs) in the JADER database. SMQ 20000041 contains 112 preferred terms (PTs) of “Hyperglycaemia/new onset diabetes mellitus.” Because SMQs are composed of MedDRA terms from one or more System Organ Classes (SOCs) to detect signals, SMQs potentially contain improper terms to define specific conditions. To exclude unrelated items, such as hypoglycemia, lipid abnormalities, or weight changes, a systematic literature search was performed with the PubMed database using terms ‘drug induced hyperglycemia’ OR ‘drug induced diabetes’ and appropriate terms were selected from those reports. Studies were restricted to those in English published between January 2000 and November 2018. Furthermore, all 112 PTs were also reviewed by our personal knowledge. Then, 51 PTs indicative of hyperglycemia or diabetes in SMQ 20000041 were selected (Table 1). We calculated the reporting odds ratios (RORs) and their 95% confidence intervals (CIs) for those 51 selected PTs indicative of hyperglycemia or diabetes in SMQ 20000041 (Table 1). Correlations are considered significant when the lower bound of the two-sided 95% confidence interval of the ROR for the risk of hyperglycemia was larger than 1. The data were analyzed using the Statistical Package for JMP 12.0.1 (SAS Institute Inc., Cary, NC, USA).
Hyperglycemia/new onset diabetes mellitus (SMQ code: 20000041) | |||||
---|---|---|---|---|---|
CODE | Preferred Term | CODE | Preferred Term | CODE | Preferred Term |
10073667 | Acquired lipoatrophic diabetes | 10018209 | Gestational diabetes | 10023392 | Ketosis-prone diabetes mellitus |
10065367 | Blood 1,5-anhydroglucitol decreased | 10018429 | Glucose tolerance impaired | 10066389 | Latent autoimmune diabetes in adults |
10005557 | Blood glucose increased | 10018430 | Glucose tolerance impaired in pregnancy | 10075980 | Monogenic diabetes |
10012596 | Diabetes complicating pregnancy | 10018478 | Glucose urine present | 10028933 | Neonatal diabetes mellitus |
10012601 | Diabetes mellitus | 10018473 | Glycosuria | 10033660 | Pancreatogenous diabetes |
10012607 | Diabetes mellitus inadequate control | 10018475 | Glycosuria during pregnancy | 10067584 | Type 1 diabetes mellitus |
10012631 | Diabetes with hyperosmolarity | 10018484 | Glycosylated haemoglobin increased | 10067585 | Type 2 diabetes mellitus |
10012650 | Diabetic coma | 10020635 | Hyperglycaemia | 10057597 | Urine ketone body present |
10012668 | Diabetic hyperglycaemic coma | 10063554 | Hyperglycaemic hyperosmolar nonketotic syndrome |
10000486 | Acidosis |
10012669 | Diabetic hyperosmolar coma | 10071394 | Hyperglycaemic seizure | 10005554 | Blood glucose abnormal |
10012671 | Diabetic ketoacidosis | 10071286 | Hyperglycaemic unconsciousness | 10049803 | Blood glucose fluctuation |
10012672 | Diabetic ketoacidotic hyperglycaemic coma |
10056997 | Impaired fasting glucose | 10018428 | Glucose tolerance decreased |
10012673 | Diabetic ketosis | 10022491 | Insulin resistant diabetes | 10018433 | Glucose tolerance test abnormal |
10074309 | Diabetic metabolic decompensation | 10053247 | Insulin-requiring type 2 diabetes mellitus | 10078582 | Indeterminate glucose tolerance |
10080061 | Euglycaemic diabetic ketoacidosis | 10023379 | Ketoacidosis | 10022494 | Insulin tolerance test abnormal |
10017395 | Fructosamine increased | 10023388 | Ketonuria | 10078891 | Urine glucose/creatinine ratio abnormal |
10072628 | Fulminant type 1 diabetes mellitus | 10023391 | Ketosis | 10078889 | Urine glucose/creatinine ratio increased |
MedDRA, Medical Dictionary for Regulatory Activities/Japanese version; SMQ, Standardized MedDRA Queries
The JADER database contained 459,250 reports of adverse drug reactions. We extracted cases of “suspected medicine”. We excluded anti-diabetes agents (58 drugs) from the data set. Table 2 shows the top 15 drugs most frequently implicated in the development of hyperglycemia or diabetes in the JADER database; prednisolone was the most frequently reported drug causing hyperglycemia (n = 698), with an ROR of 3.96 (95% CI, 3.96–4.28), followed by tacrolimus (n = 393; ROR = 3.51; 95% CI, 3.17–3.89). The mTOR inhibitor everolimus was the third most frequently reported cause of hyperglycemia or diabetes, with a higher ROR than that of the 2 drugs mentioned above (n = 340; ROR = 8.56; 95% CI, 7.65–9.57) (Table 2). We confirmed the association of antipsychotic drugs, including olanzapine (n = 270; ROR = 10.91; 95% CI, 9.62–12.38) and quetiapine (n = 247; ROR = 9.82; 95% CI, 8.61–11.20), with the risk of development of hyperglycemia. Well-known drugs for drug-induced diabetes such as ribavirin (n = 278; ROR = 1.80; 95% CI, 1.60–2.03), aripiprazole (n = 213; ROR = 7.48; 95% CI, 6.51–8.61), and atorvastatin (n = 104; ROR = 3.80; 95% CI, 3.12–4.62) were listed. Notably, a selective tyrosine kinase inhibitor nilotinib (ROR = 4.28; 95% CI, 3.48–5.27) and a human monoclonal antibody to programmed cell death 1 (PD-1), nivolumab (ROR = 3.36; 95% CI, 2.71–4.17), were also identified as potential causes of hyperglycemia.
Drug | Number of cases | Number of non-cases | Total number of reports | ROR | 95% CI |
---|---|---|---|---|---|
Prednisolone | 698 | 21,818 | 22,516 | 3.96 | 3.66–4.28 |
Tacrolimus Hydrate | 393 | 13,511 | 13,904 | 3.51 | 3.17–3.89 |
Everolimus | 340 | 4,806 | 5,146 | 8.56 | 7.65–9.57 |
Ribavirin | 278 | 18,327 | 18,605 | 1.80 | 1.60–2.03 |
Olanzapine | 270 | 2,975 | 3,245 | 10.91 | 9.62–12.38 |
Quetiapine Fumarate | 247 | 3,017 | 3,264 | 9.82 | 8.61–11.20 |
Aripiprazole | 213 | 3,402 | 3,615 | 7.48 | 6.51–8.61 |
Interferon Alfa-2b (Genetical Recombination) | 203 | 12,598 | 12,801 | 1.91 | 1.66–2.20 |
Risperidone | 124 | 4,647 | 4,771 | 3.16 | 2.64–3.78 |
Atorvastatin Calcium Hydrate | 104 | 3,237 | 3,341 | 3.80 | 3.12–4.62 |
Dexamethasone | 99 | 6,087 | 6,186 | 1.92 | 1.57–2.34 |
Ciclosporin | 93 | 11,520 | 11,613 | 0.95 | 0.77–1.16 |
Nilotinib Hydrochloride Hydrate | 93 | 2,565 | 2,658 | 4.28 | 3.48–5.27 |
Methylprednisolone Sodium Succinate | 91 | 2,834 | 2,925 | 3.79 | 3.07–4.68 |
Nivolumab (Genetical Recombination) | 86 | 3,017 | 3,103 | 3.36 | 2.71–4.17 |
ROR, reporting odds ratio. 95% CI, 95% confidence interval.
Hyperglycemia or diabetes is a widely encountered adverse effect of prednisolone and tacrolimus, often seen in patients with autoimmune diseases and post-transplant patients. However, the prevalence of hyperglycemia caused by everolimus has been little known compared to the above 2 drugs in Japan, probably because the mTOR inhibitors are relatively recently introduced drugs to clinical practice. In the present study, we showed that everolimus is the third most frequently reported cause of drug-induced hyperglycemia or diabetes in Japanese patients in the JADER database, with a higher ROR than that of prednisolone or tacrolimus. Hyperglycemia induced by mTOR inhibitors was reported to occur at an incidence ranging from 13% to 50% in clinical trials in which they were used as anticancer drugs [1, 2, 4, 5, 12-16]. Hyperglycemia associated with mTOR inhibitors is thought to be caused by both impaired insulin secretion and insulin resistance. mTOR inhibition by AZD8055, a selective mTOR kinase inhibitor, has been shown to reduce insulin-stimulated glucose uptake in muscles [7], whereas mTOR inhibition caused by rapamycin was shown to be associated with reduced glucose-induced insulin secretion from the pancreatic beta cells in vivo [8]. Since Japanese non-obese subjects with diabetes have impaired insulin secretion rather than insulin resistance in the early phase, the impact of everolimus on the pancreatic beta cells could be a critical issue.
The results of this study consistently showed the same set of drugs as inducing hyperglycemia, and ROR is now a well-established pharmacovigilance index [17]. Spontaneous reporting systems, including the JADER database, however, have limitations in terms of inherent biases and a lack of data on controls. The biases include underreporting, overreporting, missing data, input, selection, and many kinds of biases, which affect the RORs. Therefore, ROR does not reflect exact prevalence rate and it might be unsuitable for comparing risks between different drugs, which must always be taken into consideration. The contribution of coexisting illnesses, drug dose, or the period of exposure were not investigated in this study. Studies in cohorts of Japanese patients receiving everolimus are required for accurate clarification of the prevalence. We recommend close monitoring of post-transplant patients and cancer patients receiving everolimus to prevent the complication of hyperglycemia, as these patients are more prone to developing diabetes.
The authors acknowledge all the contributors of JADER database. The authors thank Misa Katayama (Yokohama City University) for her excellent secretarial assistance. This work was partly supported by Grant-in-Aid for Young Scientists (B) 18K16240 from MEXT of Japan (to J.S.) and Grants-in-Aid for Scientific Research (B) 16H05329 from MEXT of Japan (to Y.T.).
None of the authors have any potential conflicts of interest associated with this research.