Abstract
Case reports and a pharmacovigilance analysis have linked glucagon-like peptide 1 receptor agonists (GLP-1 RAs) with anaphylactic reactions, but real-world evidence for this possible association is lacking. Using databases from the United Kingdom (Clinical Practice Research Datalink) and the United States (Medicare, Optum (Optum, Inc., Eden Prairie, Minnesota), and IBM MarketScan (IBM, Armonk, New York)), we employed a new-user, active comparator study design wherein initiators of GLP-1 RAs were compared with 2 different active comparator groups (initiators of dipeptidyl peptidase 4 (DPP-4) inhibitors and initiators of sodium-glucose cotransporter 2 (SGLT-2) inhibitors) between 2007 and 2019. Propensity score fine stratification weighted Cox proportional hazards models were fitted to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for an anaphylactic reaction. Database-specific HRs were pooled using random-effects models. Compared with the use of DPP-4 inhibitors (n = 1,641,520), use of GLP-1 RAs (n = 324,098) generated a modest increase in the HR for anaphylactic reaction, with a wide 95% CI (36.9 per 100,000 person-years vs. 32.1 per 100,000 person-years, respectively; HR = 1.15, 95% CI: 0.94, 1.42). Compared with SGLT-2 inhibitors (n = 366,067), GLP-1 RAs (n = 259,929) were associated with a 38% increased risk of anaphylactic reaction (40.7 per 100,000 person-years vs. 29.4 per 100,000 person-years, respectively; HR = 1.38, 95% CI: 1.02, 1.87). In this large, multisite population-based cohort study, GLP-1 RAs were associated with a modestly increased risk of anaphylactic reaction when compared with DPP-4 inhibitors and SGLT-2 inhibitors.
Keywords: anaphylactic reaction, dipeptidyl peptidase 4 inhibitors, glucagon-like peptide 1 receptor agonists, pharmacoepidemiology, sodium-glucose cotransporter 2 inhibitors
Abbreviation
- CI
confidence interval
- CPRD
Clinical Practice Research Datalink
- DPP-4
dipeptidyl peptidase 4
- GLP-1 RA
glucagon-like peptide 1 receptor agonist
- HR
hazard ratio
- ICD-10
International Classification of Diseases, Tenth Revision
- SGLT-2
sodium-glucose cotransporter 2
Editor’s note: An invited commentary on this article appears on page 1368, and the authors’ response appears on page 1372.
Glucagon-like peptide 1 receptor agonists (GLP-1 RAs) are effective second- to third-line drugs used in the treatment of type 2 diabetes. These drugs reduce body weight and have beneficial effects on cardiovascular and renal outcomes (1, 2). However, there are concerns that their use might be associated with an anaphylactic reaction, a rare but potentially life-threatening adverse event (3).
GLP-1 RAs enhance glucose-dependent insulin release by mimicking the actions of the gut hormone glucagon-like peptide 1 (4). However, GLP-1 RAs have structural differences with endogenous glucagon-like peptide 1; these have been shown to elicit immunogenicity, resulting in antidrug antibodies in up to 70% of recipients (3, 5). While the clinical significance of these antibodies is unclear, anaphylactic reactions have been reported in randomized controlled trials of GLP-1 RAs (6). Indeed, the relatively high frequency of this adverse event halted the clinical development of a GLP-1 RA, taspoglutide (7). Over the years, several case reports and a pharmacovigilance analysis have documented anaphylactic reactions among patients using GLP-1 RAs, but their interpretation has been limited by potential reporting bias (8–13). Interestingly, in one case report, an anaphylactic reaction occurred after 10 months of continuous GLP-1 RA use, (9) while in other case reports, the event occurred after the reintroduction of GLP-1 RAs in patients with a history of use of this drug (10–12). To date, no real-world studies have been conducted to assess the risk of anaphylactic reaction associated with the use of GLP-1 RAs as compared with other second- to third-line antidiabetic drugs. This is relevant in the context of comparative safety, given that most antidiabetic drugs have been linked to hypersensitivity reactions (14, 15).
Given the increasing use of GLP-1 RAs (16) and the fact that drug-induced anaphylactic reactions can be fatal (17), we conducted a large, multisite population-based cohort study to determine whether GLP-1 RAs are associated with an increased risk of anaphylactic reaction when compared with dipeptidyl peptidase 4 (DPP-4) inhibitors and sodium-glucose cotransporter 2 (SGLT-2) inhibitors among patients with type 2 diabetes.
METHODS
Data sources
This was a multisite cohort study using 4 population-based databases, including 1 from the United Kingdom (Clinical Practice Research Datalink (CPRD), including both the GP OnLine Data (GOLD) and Aurum databases), and 3 from the United States (Medicare fee-for-service data, Optum Clinformatics Data Mart (Optum, Inc., Eden Prairie, Minnesota), and IBM MarketScan Research Databases (IBM, Armonk, New York)). The CPRD is a general practice database maintained by the UK Department of Health and Social Care containing detailed medical diagnoses and records of written prescriptions for over 50 million patients seen at more than 2,000 general practices in the United Kingdom. In addition, the CPRD was linked to the UK National Health Service’s Hospital Episode Statistics repository and the UK Office for National Statistics. The Medicare database, maintained by the US Department of Health and Human Services, contains outpatient and inpatient health-care usage data and records of filled prescriptions for over 55 million federally covered patients in the United States. Finally, Optum Clinformatics Data Mart and IBM MarketScan are 2 commercial databases that capture data on outpatient and inpatient health-care use and records of filled prescriptions for over 60 million patients in the United States. The study protocol was approved by the Independent Scientific Advisory Committee of the CPRD, the Research Ethics Board of the Jewish General Hospital (Montreal, Quebec, Canada), and the Brigham and Women’s Hospital Institutional Review Board (Boston, Massachusetts).
Study population
We employed a new-user, active comparator study design wherein initiators of GLP-1 RAs were compared with 2 sets of comparators: initiators of DPP-4 inhibitors (primary comparator group) and initiators of SGLT-2 inhibitors (secondary comparator group). Users of DPP-4 inhibitors were selected as the primary comparator group because DPP-4 inhibitors are also incretin-based drugs, they are used at the same disease stage, and they were introduced in the same year as GLP-1 RAs (2). However, because there were concerns that DPP-4 inhibitors may be associated with severe hypersensitivity reactions, including anaphylactic reactions (14, 18), we selected users of SGLT-2 inhibitors as a secondary comparator group; these drugs have not been associated with severe allergic reactions and are used at a similar disease stage as GLP-1 RAs (2). Thus, for the primary comparison, we identified patients with a new prescription for a GLP-1 RA or a DPP-4 inhibitor from January 1, 2007 (the year when both GLP-1 RAs and DPP-4 inhibitors became available) to the end of data availability (December 31, 2017, for Medicare; November 30, 2018, for the CPRD; December 31, 2018, for MarketScan; and December 31, 2019, for Optum). For the secondary comparison, we identified patients initiating use of a GLP-1 RA or an SGLT-2 inhibitor between March 1, 2013 (when SGLT-2 inhibitors became available) and the end of data availability. For both analyses, cohort entry was defined as the first prescription of either the GLP-1 RA or a comparator drug during the study period.
To be included in the respective exposure groups, all patients were required to be at least 18 years of age (or at least 66 years of age in Medicare), have at least 1 year of medical history and continuous enrollment in each database, and have at least 1 diagnosis of type 2 diabetes during the year before cohort entry. We excluded patients who were prescribed the exposure drugs in the year before cohort entry, including the 3-mg formulation of liraglutide, and those who were prescribed more than 1 type of the exposure drugs at cohort entry. We also excluded patients with type 1 diabetes, patients with idiopathic intracranial hypertension (in the CPRD, since this is an alternative indication for exenatide use in European Union countries), and those prescribed insulin (since these represent patients with more advanced diabetes, and insulin is associated with anaphylactic reactions) (19) during the year before and including cohort entry. Finally, we excluded patients diagnosed in the year before and including cohort entry with rare conditions that may affect the immune system, including human immunodeficiency virus/acquired immunodeficiency syndrome and end-stage renal disease, and organ transplant recipients. Web Figure 1 (available at https://doi.org/10.1093/aje/kwac021) shows the study design.
Follow-up period
We used an on-treatment exposure definition where patients were followed from cohort entry to an anaphylactic reaction (defined below), treatment discontinuation or crossover to one of the study drugs, death, the end of registration or enrollment, or the end of the study period, whichever occurred first. Patients were considered continuously exposed if the duration of 1 prescription plus a 90-day grace period overlapped with the date of the subsequent prescription. Remaining prescription durations were ignored and not carried over to the subsequent prescription. Treatment discontinuation was defined by the absence of a refill prescription by the end of the 90-day grace period. This grace period was chosen to accommodate the fact that peak antibody levels with GLP-1 RAs occur 12–14 weeks after treatment initiation (20) and that anaphylactic reactions might result from the reintroduction of an agent after temporary discontinuation (10–12). GLP-1 RAs are available as daily or weekly subcutaneous injections and DPP-4 inhibitors and SGLT-2 inhibitors as daily tablets in monthly refills.
Outcome definition
In the CPRD, anaphylactic reactions were identified using outpatient data (CPRD; Read codes), inpatient data (Hospital Episode Statistics; International Classification of Diseases, Tenth Revision (ICD-10) codes), and vital statistics (Office for National Statistics; ICD-10 codes). This definition has been shown to have a positive predictive value of 72.5% (21). If patients had events recorded in more than 1 database, the earliest event was set as the date of the outcome. For the Medicare, Optum, and MarketScan databases, we adapted a validated outcome definition for anaphylactic reaction that has been shown to have a positive predictive value of 63% (22) based on events recorded in the inpatient, emergency department, or outpatient setting (22). Overall, these outcome definitions have previously been used to answer drug or vaccine safety questions focusing on anaphylactic reactions in UK- and US-based studies (23–25). The Read codes and International Classification of Diseases, Ninth Revision/ICD-10 codes used to define anaphylactic reaction in the different databases are listed in Web Table 1.
Potential confounders
We considered a total of 93 potential confounders. These included the following variables, measured at cohort entry or during the year before cohort entry: age, sex, proxies for socioeconomic status (in US databases), and smoking status (CPRD). We also considered factors that can influence immune status, including obesity, alcohol-related disorders, drug abuse, cancer (other than nonmelanoma skin cancer), and chronic kidney disease. We considered the use of antiallergic drugs, as well as a history of allergic conditions, including previous anaphylactic reactions, angioedema, asthma, drug allergies, eczema, food allergies, chronic conjunctivitis, chronic rhinitis, urticaria, Hymenoptera (bee) venom allergy, and others (including Steven-Johnson syndrome, toxic epidermal necrolysis, and mastocytosis). We included variables related to diabetes severity, including hemoglobin A1c (available in CPRD), types of antidiabetic drugs used in the year before cohort entry, and microvascular and macrovascular complications of diabetes recorded in the year before cohort entry. Finally, we accounted for common comorbid conditions, commonly ordered metabolic biochemical tests (in US databases), other commonly used prescription drugs, and markers of health-care utilization. A complete list of the covariates can be found in Web Table 2.
Statistical analysis
We used propensity score fine stratification to control for confounding (26). We estimated the predicted probability of receiving a GLP-1 RA versus a comparator drug (DPP-4 inhibitor or SGLT-2 inhibitor) using multivariable logistic regression models conditional on the potential confounders listed above. Patients in the nonoverlapping regions of the propensity scores were trimmed, and 50 strata based on the propensity score distribution of the GLP-1 RA patients were created. Within each stratum, patients in the GLP-1 RA group received a weight of 1, while patients in the comparator group were reweighted proportional to the number of exposed individuals in the stratum (26).
Descriptive statistics were used to summarize the characteristics of the exposure groups before and after propensity score fine stratification weighting. Covariate balance between the comparator groups was examined using standardized differences, with standardized differences less than 0.10 being considered indicative of good balance. Weighted incidence rates of anaphylactic reaction, with 95% confidence intervals (CIs) based on the negative binomial distribution, were calculated for each exposure group. We also plotted weighted Kaplan-Meier curves to visualize the cumulative incidence of anaphylactic reaction for each exposure group over the follow-up period. Finally, weighted Cox proportional hazards models were fitted to estimate hazard ratios (HRs) and 95% CIs for an anaphylactic reaction associated with use of GLP-1 RAs versus DPP-4 inhibitors (primary comparison) and SGLT-2 inhibitors (secondary comparison).
Secondary analyses.
We performed 4 secondary analyses based on the primary comparison (GLP-1 RAs vs. DPP-4 inhibitors). The first 2 assessed whether there was effect modification by age (<65 years vs. ≥65 years) and history of allergic conditions (a composite of anaphylactic reactions, angioedema, asthma, drug allergies, eczema, food allergy, chronic conjunctivitis, chronic rhinitis, urticaria, Hymenoptera venom allergy, and other allergic disorders, including Steven-Johnson syndrome, toxic epidermal necrolysis, and mastocytosis). Effect modification was assessed by including an interaction term for interaction between the exposures and these variables. Third, we stratified by type of GLP-1 RA, given that a pharmacovigilance analysis observed higher reporting odds of anaphylactic reaction with exendin-based GLP-1 RAs (exenatide and lixisenatide) than with human analog GLP-1 RAs (liraglutide, dulaglutide, albiglutide, and semaglutide) (13). Finally, because Ornelas et al. (9) reported anaphylactic reactions after a delayed exposure to GLP-1 RAs, we conducted 2 analyses to examine the effect of duration of use. First, we stratified the follow-up period into categories: ≤30 days, 31–365 days, and >365 days. Second, we restricted the analysis to anaphylactic reactions occurring after 1 year of treatment initiation among the study participants who were still in treatment and uncensored at that point in time.
Sensitivity analyses.
We conducted 6 sensitivity analyses based on the primary comparison (GLP-1 RAs vs. DPP-4 inhibitors). First, we repeated the primary analyses using 2 alternative grace periods of 60 days and 120 days between nonsuccessive prescriptions. Second, we used an intention-to-treat exposure definition to account for potential informative censoring and the fact that anaphylactic reactions may be delayed because of long-term and potentially irreversible maintenance of immunoglobulin E–mediated immune memory; this may lead to precipitation of anaphylactic reactions upon reintroduction of the study drugs or related allergens, even after treatment termination (27). Third, we excluded patients with a history of cancer to account for the fact that overall immune status might be altered in these patients. Fourth, since insulin has also been associated with anaphylactic reactions (19), we repeated the primary analyses by additionally censoring on insulin initiation during the follow-up period. Fifth, because other commonly used drug classes, such as nonsteroidal antiinflammatory drugs and antibiotics, have also been associated with anaphylactic reactions, we conducted an analysis additionally censoring on initiation of these 2 drug classes during the follow-up period (28). Finally, to increase the sensitivity of the outcome definition, we modified the definition to include severe allergic reactions recorded both in the inpatient setting and at emergency departments (but requiring intervention with diphenhydramine or epinephrine).
Pooling database-specific estimates.
The analyses were conducted separately in each of the 4 databases, and pooling was conducted using a random-effect meta-analytical model with inverse variance weighting. Between-database heterogeneity was assessed using the I2 statistic. All analyses were conducted with SAS, version 9.4 (SAS Institute, Inc., Cary, North Carolina), Aetion Evidence Platform (Aetion, Inc., New York, New York), and STATA, version 14 (StataCorp LLC, College Station, Texas).
RESULTS
For the primary comparison, 324,098 and 1,641,520 patients newly treated with GLP-1 RAs and DPP-4 inhibitors met the study inclusion criteria, respectively (Web Figure 2). Overall, there were 851 anaphylactic reactions during a total of 2,597,396 person-years of follow-up, yielding an overall incidence rate of 32.8 per 100,000 person-years (95% CI: 30.6, 35.0).
Before propensity score fine stratification weighting, GLP-1 RA users were younger than DPP-4 inhibitor users, more likely to be female, and more likely to be obese, while they were less likely than DPP-4 inhibitor users to have macrovascular complications (Table 1). The exposure groups had similar histories of allergic conditions and use of antiallergic drugs. After propensity score fine stratification weighting, the characteristics were well-balanced between the exposure groups (database-specific standardized differences are presented in Web Table 3).
Table 1.
Baseline Characteristics of Initiators of Glucagon-Like-Peptide 1 Receptor Agonists and Dipeptidyl Peptidase 4 Inhibitors Before and After Propensity Score Fine Stratification Weighting in 4 UK and US Databases, 2007–2019
| Before Weighting | After Weighting | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
|
GLP-1 RAs
(n = 328,096) |
DPP-4 Inhibitors
a
(n = 1,661,620) |
ASD |
GLP-1 RAs
(n = 324,098) |
DPP-4 Inhibitors
a
(n = 1,641,520) |
ASD | |||||
| Characteristic | No. | % | No. | % | No. | % | No. | % | ||
| Age, yearsb | 58.1 (9.9) | 63.3 (10.4) | 0.50 | 58.2 (9.9) | 58 (9.2) | 0.02 | ||||
| Female sex | 181,864 | 55.4 | 818,199 | 49.2 | 0.12 | 179,907 | 55.5 | 917,704 | 55.9 | 0.01 |
| Ever smoking | 43,955 | 13.4 | 177,663 | 10.7 | 0.08 | 43,751 | 13.5 | 194,620 | 11.9 | 0.05 |
| Alcohol dependence | 2,956 | 0.9 | 15,552 | 0.9 | 0.00 | 2,940 | 0.9 | 12,168 | 0.7 | 0.02 |
| Drug dependence | 4,856 | 1.5 | 16,121 | 1.0 | 0.05 | 4,834 | 1.5 | 20,678 | 1.3 | 0.02 |
| Obesity | 110,165 | 33.6 | 267,345 | 16.1 | 0.41 | 109,233 | 33.7 | 503,214 | 30.7 | 0.07 |
| Cancer | 30,915 | 9.4 | 209,986 | 12.6 | 0.10 | 30,696 | 9.5 | 168,525 | 10.3 | 0.03 |
| Chemotherapy | 9,525 | 2.9 | 61,307 | 3.7 | 0.04 | 9,493 | 2.9 | 50,181 | 3.1 | 0.01 |
| Chronic kidney disease | 27,518 | 8.4 | 203,264 | 12.2 | 0.13 | 27,325 | 8.4 | 149,982 | 9.1 | 0.02 |
| Anaphylactic reaction | 431 | 0.1 | 1,441 | 0.1 | 0.01 | 430 | 0.1 | 2,001 | 0.1 | 0.00 |
| Angioedema | 967 | 0.3 | 4,490 | 0.3 | 0.00 | 965 | 0.3 | 4,820 | 0.3 | 0.00 |
| Stevens-Johnson syndrome and TEN | 22 | 0.0 | 116 | 0.0 | 0.00 | 22 | 0.0 | 107 | 0.0 | 0.00 |
| Mastocytosis | 16 | 0.0 | 29 | 0.0 | 0.01 | 16 | 0.0 | 47 | 0.0 | 0.00 |
| Asthma | 31,351 | 9.6 | 122,566 | 7.4 | 0.08 | 31,201 | 9.6 | 152,703 | 9.3 | 0.01 |
| Drug allergies | 15,599 | 4.8 | 68,662 | 4.1 | 0.03 | 15,543 | 4.8 | 73,021 | 4.4 | 0.02 |
| Eczema | 17,072 | 5.2 | 95,678 | 5.8 | 0.02 | 16,979 | 5.2 | 87,057 | 5.3 | 0.00 |
| Food allergy | 3,517 | 1.1 | 18,148 | 1.1 | 0.00 | 3,496 | 1.1 | 17,978 | 1.1 | 0.00 |
| Allergic rhinitis | 33,234 | 10.1 | 143,525 | 8.6 | 0.05 | 33,020 | 10.2 | 170,429 | 10.4 | 0.01 |
| Allergic conjunctivitis | 3,045 | 0.9 | 17,463 | 1.1 | 0.01 | 3,028 | 0.9 | 16,347 | 1.0 | 0.01 |
| Urticaria | 3,394 | 1.0 | 13,725 | 0.8 | 0.02 | 3,375 | 1.0 | 15,231 | 0.9 | 0.01 |
| Hymenoptera venom allergy | 318 | 0.1 | 650 | 0.0 | 0.02 | 317 | 0.1 | 885 | 0.1 | 0.02 |
| Antihistamines | ||||||||||
| Nonoral | 16,281 | 5.0 | 68,712 | 4.1 | 0.04 | 16,227 | 5.0 | 84,909 | 5.2 | 0.01 |
| Oral | 27,572 | 8.4 | 125,423 | 7.6 | 0.03 | 27,572 | 8.5 | 136,798 | 8.3 | 0.01 |
| Corticosteroids | ||||||||||
| Oral | 56,943 | 17.4 | 245,918 | 14.8 | 0.07 | 56,942 | 17.6 | 291,627 | 17.8 | 0.01 |
| Orally inhaled | 22,880 | 7.0 | 106,863 | 6.4 | 0.02 | 22,879 | 7.1 | 117,713 | 7.2 | 0.00 |
| Epinephrine injection | 4,081 | 1.2 | 15,724 | 1.0 | 0.03 | 4,077 | 1.3 | 22,109 | 1.3 | 0.01 |
| Biologics | 9,286 | 2.8 | 57,137 | 3.4 | 0.03 | 9,239 | 2.9 | 50,565 | 3.1 | 0.01 |
| Stable angina | 11,455 | 3.5 | 70,211 | 4.2 | 0.04 | 11,395 | 3.5 | 60,550 | 3.7 | 0.01 |
| Unstable angina | 6,573 | 2.0 | 42,916 | 2.6 | 0.04 | 6,546 | 2.0 | 32,276 | 2.0 | 0.00 |
| Myocardial infarction | 20,346 | 6.2 | 172,879 | 10.4 | 0.08 | 10,180 | 3.1 | 55,419 | 3.4 | 0.01 |
| Other coronary atherosclerotic diseases | 53,052 | 16.2 | 373,533 | 22.5 | 0.16 | 52,733 | 16.3 | 302,339 | 18.4 | 0.06 |
| Heart failure | 1,386 | 0.4 | 7,615 | 0.5 | 0.15 | 20,246 | 6.2 | 117,975 | 7.2 | 0.04 |
| Previous cardiac procedure | 3,167 | 1.0 | 25,185 | 1.5 | 0.05 | 3,151 | 1.0 | 14,504 | 0.9 | 0.01 |
| Peripheral circulatory disorder | 50,162 | 15.3 | 234,189 | 14.1 | 0.13 | 33,254 | 10.3 | 187,369 | 11.4 | 0.04 |
| Stroke | 17,773 | 5.4 | 152,737 | 9.2 | 0.15 | 17,658 | 5.4 | 103,330 | 6.3 | 0.04 |
| Transient ischemic attack | 4,889 | 1.5 | 41,208 | 2.5 | 0.07 | 4,863 | 1.5 | 27,066 | 1.6 | 0.01 |
| Neuropathy | 10,230 | 3.1 | 78,458 | 4.7 | 0.03 | 49,733 | 15.3 | 261,142 | 15.9 | 0.02 |
| Gastroparesis | 20,008 | 6.1 | 107,947 | 6.5 | 0.01 | 1,374 | 0.4 | 8,110 | 0.5 | 0.01 |
| Retinopathy | 38,051 | 11.6 | 247,939 | 14.9 | 0.02 | 19,773 | 6.1 | 94,897 | 5.8 | 0.01 |
| Nephropathy | 33,480 | 10.2 | 241,238 | 14.5 | 0.10 | 24,389 | 7.5 | 117,187 | 7.1 | 0.01 |
| Hemoglobin A1c concentration, %c | ||||||||||
| ≤7.0 | 1,085 | 9.4 | 9,561 | 8.6 | 0.03 | 1,084 | 9.4 | 10,871 | 9.8 | 0.01 |
| 7.1–8.0 | 2,047 | 17.8 | 32,854 | 29.4 | 0.28 | 2,047 | 17.8 | 19,957 | 18.0 | 0.00 |
| >8.0 | 8,309 | 72.2 | 68,592 | 61.4 | 0.23 | 8,309 | 72.2 | 79,710 | 71.7 | 0.01 |
| No. of antidiabetic drugs at cohort entryb | 0.9 (0.7) | 0.9 (0.7) | 0.10 | 0.9 (0.7) | 0.9 (0.7) | 0.06 | ||||
| Metformin | ||||||||||
| Current use | 113,445 | 34.6 | 472,301 | 28.4 | 0.13 | 113,445 | 35.0 | 574,785 | 35.0 | 0.00 |
| Past use | 238,701 | 72.8 | 1,117,844 | 67.3 | 0.12 | 238,699 | 73.7 | 1,203,991 | 73.3 | 0.01 |
| Sulfonylureas | ||||||||||
| Current use | 59,840 | 18.2 | 343,199 | 20.7 | 0.06 | 59,840 | 18.5 | 319,140 | 19.4 | 0.02 |
| Past use | 135,714 | 41.4 | 729,823 | 43.9 | 0.05 | 135,713 | 41.9 | 706,363 | 43.0 | 0.02 |
| Thiazolidinediones | ||||||||||
| Current use | 21,062 | 6.4 | 95,083 | 5.7 | 0.03 | 21,062 | 6.5 | 111,057 | 6.8 | 0.01 |
| Past use | 56,315 | 17.2 | 299,547 | 18.0 | 0.02 | 56,313 | 17.4 | 291,588 | 17.8 | 0.01 |
| SGLT-2 inhibitors | ||||||||||
| Current use | 12,281 | 3.7 | 14,836 | 0.9 | 0.19 | 12,281 | 3.8 | 50,152 | 3.1 | 0.04 |
| Past use | 25,454 | 7.8 | 33,675 | 2.0 | 0.27 | 25,452 | 7.9 | 108,324 | 6.6 | 0.05 |
| α-Glucosidase inhibitors (ever use) | 1,632 | 0.5 | 8,886 | 0.5 | 0.01 | 1,632 | 0.5 | 8,927 | 0.5 | 0.01 |
| Meglitinides (ever use) | 4,790 | 1.5 | 32,050 | 1.9 | 0.04 | 4,790 | 1.5 | 27,357 | 1.7 | 0.02 |
| Obstructive sleep apnea | 55,335 | 16.9 | 145,604 | 8.8 | 0.24 | 54,962 | 17.0 | 277,372 | 16.9 | 0.00 |
| Osteoarthritis | 67,713 | 20.6 | 353,536 | 21.3 | 0.02 | 67,319 | 20.8 | 366,905 | 22.4 | 0.04 |
| Other arthropathies | 152,733 | 46.6 | 763,548 | 46.0 | 0.01 | 151,582 | 46.8 | 803,121 | 48.9 | 0.04 |
| Dorsopathies | 103,421 | 31.5 | 470,982 | 28.3 | 0.07 | 102,702 | 31.7 | 527,997 | 32.2 | 0.01 |
| Pancreatitis | 1,416 | 0.4 | 11,156 | 0.7 | 0.03 | 1,408 | 0.4 | 7,137 | 0.4 | 0.00 |
| Urinary tract infections | 44,643 | 13.6 | 255,641 | 15.4 | 0.05 | 44,363 | 13.7 | 229,755 | 14.0 | 0.01 |
| Delirium or psychosis | 4,996 | 1.5 | 53,823 | 3.2 | 0.11 | 4,977 | 1.5 | 27,549 | 1.7 | 0.01 |
| Dementia | 5,870 | 1.8 | 87,774 | 5.3 | 0.19 | 5,834 | 1.8 | 35,717 | 2.2 | 0.03 |
| Angiotensin-converting enzyme inhibitors | 141,225 | 43.0 | 741,690 | 44.6 | 0.03 | 141,225 | 43.6 | 720,889 | 43.9 | 0.01 |
| Angiotensin II receptor blockers | 90,662 | 27.6 | 463,430 | 27.9 | 0.01 | 90,661 | 28.0 | 478,382 | 29.1 | 0.03 |
| Beta blockers | 101,185 | 30.8 | 617,105 | 37.1 | 0.13 | 101,184 | 31.2 | 551,405 | 33.6 | 0.05 |
| Calcium channel blockers | 76,853 | 23.4 | 465,809 | 28.0 | 0.11 | 76,853 | 23.7 | 406,131 | 24.7 | 0.02 |
| Diuretics | 93,376 | 28.5 | 490,020 | 29.5 | 0.02 | 93,375 | 28.8 | 507,467 | 30.9 | 0.05 |
| Other antihypertensive drugs | 17,213 | 5.3 | 112,019 | 6.7 | 0.06 | 17,213 | 5.3 | 94,954 | 5.8 | 0.02 |
| Antiplatelet agents | 30,725 | 9.4 | 196,638 | 11.8 | 0.08 | 30,725 | 9.5 | 154,232 | 9.4 | 0.00 |
| Statins | 203,361 | 62.0 | 1,062,898 | 64.0 | 0.04 | 203,360 | 62.7 | 1,046,138 | 63.7 | 0.02 |
| Warfarin | 11,173 | 3.4 | 90,555 | 5.5 | 0.10 | 11,173 | 3.4 | 65,489 | 4.0 | 0.03 |
| Oral anticoagulants (except warfarin) | 6,610 | 2.0 | 32,147 | 1.9 | 0.01 | 6,610 | 2.0 | 36,024 | 2.2 | 0.01 |
| Injectable anticoagulants | 2,417 | 0.7 | 13,594 | 0.8 | 0.01 | 2,417 | 0.7 | 13,364 | 0.8 | 0.01 |
| Nonsteroidal antiinflammatory drugs | 84,922 | 25.9 | 375,705 | 22.6 | 0.08 | 84,920 | 26.2 | 428,987 | 26.1 | 0.00 |
| Antibiotic use | ||||||||||
| Past use | 185,169 | 56.4 | 886,329 | 53.3 | 0.06 | 185,167 | 57.1 | 951,457 | 58.0 | 0.02 |
| Recent use | 37,724 | 11.5 | 191,677 | 11.5 | 0.00 | 37,722 | 11.6 | 193,161 | 11.8 | 0.00 |
| Nonbiological DMARDs | 8,134 | 2.5 | 41,921 | 2.5 | 0.00 | 8,134 | 2.5 | 43,720 | 2.7 | 0.01 |
| Bone mineral density | 19,500 | 5.9 | 103,349 | 6.2 | 0.01 | 19,363 | 6.0 | 108,953 | 6.6 | 0.03 |
| Colonoscopy | 32,426 | 9.9 | 159,191 | 9.6 | 0.01 | 32,223 | 9.9 | 167,521 | 10.2 | 0.01 |
| Influenza vaccine | 114,504 | 34.9 | 594,583 | 35.8 | 0.02 | 113,735 | 35.1 | 627,282 | 38.2 | 0.06 |
| Mammogram | 82,052 | 25.0 | 338,572 | 20.4 | 0.11 | 81,410 | 25.1 | 428,567 | 26.1 | 0.02 |
| Papanicolaou smear | 40,598 | 12.4 | 161,919 | 9.7 | 0.08 | 40,331 | 12.4 | 205,935 | 12.5 | 0.00 |
| Pneumococcal vaccine | 67,120 | 20.5 | 237,590 | 14.3 | 0.16 | 66,552 | 20.5 | 346,844 | 21.1 | 0.01 |
| Prostate examination (DRE or PSA test) | 25,266 | 7.8 | 135,770 | 8.3 | 0.02 | 25,061 | 7.7 | 135,232 | 8.2 | 0.02 |
| Hospitalization | ||||||||||
| Recent (≤30 days prior to cohort entry) | 32,702 | 10.1 | 269,545 | 16.4 | 0.19 | 32,524 | 10.0 | 180,116 | 11.0 | 0.03 |
| Older (31–365 days prior to cohort entry) | 38,803 | 12.0 | 260,926 | 15.9 | 0.11 | 38,510 | 11.9 | 193,708 | 11.8 | 0.00 |
| Basic or comprehensive metabolic blood chemistry testd | 254,277 | 81.3 | 1,224,041 | 80.0 | 0.04 | 251,931 | 80.6 | 1,257,618 | 82.2 | 0.04 |
| Frailty scoreb,d | 0.1 (0.0) | 0.2 (0.1) | 1.41 | 0.1 (0.0) | 0.1 (0.1) | 0.00 | ||||
| No. of cardiologist visitsb,d | 2 (5.7) | 2.4 (8.0) | 0.06 | 2 (5.7) | 2 (6.5) | 0.00 | ||||
| No. of distinct drug prescriptionsb,d | 11.2 (6.2) | 10.3 (6.0) | 0.15 | 11.3 (6.1) | 11.3 (6.4) | 0.00 | ||||
| No. of emergency department visitsb,d | 0.5 (1.9) | 0.6 (2.1) | 0.05 | 0.5 (1.9) | 0.5 (1.7) | 0.00 | ||||
| No. of endocrinologist visitsb,d | 1.4 (5.7) | 0.6 (4.2) | 0.16 | 1.4 (5.7) | 1.3 (7.3) | 0.02 | ||||
| No. of hemoglobin A1c tests orderedb,d | 2.2 (1.3) | 2 (1.4) | 0.15 | 2.2 (1.3) | 2.2 (1.4) | 0.00 | ||||
| No. of internal medicine/family medicine visitsb,d | 12.8 (17.1) | 13.7 (19.8) | 0.05 | 12.9 (17.2) | 12.9 (17.8) | 0.00 | ||||
| SES proxy measure: pharmacy copay (US dollars)b,d | 23.6 (27.8) | 23 (29.2) | 0.02 | 23.6 (27.8) | 23.6 (35.4) | 0.00 | ||||
Abbreviations: ASD, absolute standardized difference; DMARD, disease-modifying antirheumatic drug; DPP-4, dipeptidyl peptidase 4; DRE, digital rectal examination; GLP-1 RA, glucagon-like peptide receptor agonist; PSA, prostate-specific antigen; SD, standard deviation; SES, socioeconomic status; SGLT-2, sodium glucose cotransporter 2; TEN, toxic epidermal necrolysis.
a Pooling was conducted by taking the average of the mean values and proportions of the exposed group and taking a weighted average of the mean values and proportions of the unexposed group, the weight being the ratio of the number of exposed individuals in that database to the number of all exposed individuals.
b Values are expressed as mean (standard deviation).
c This information was available only in the Clinical Practice Research Datalink.
d This information was available only in the Medicare, Optum (Optum, Inc., Eden Prairie, Minnesota), and IBM MarketScan (IBM, Armonk, New York) databases.
Compared with DPP-4 inhibitors, the HR for anaphylactic reaction with GLP-1 RAs was slightly elevated, but with a 95% CI that included the null value (36.9/100,000 person-years vs. 32.1/100,000 person-years; HR = 1.15, 95% CI: 0.94, 1.42) (Table 2). Overall, the cumulative incidence curves did not significantly diverge, except in the CPRD, where the curves diverged after 4 years of use (Web Figure 3). The secondary analyses are summarized in Web Figure 4. There was no effect-measure modification by age or history of allergic conditions, although an increased risk was observed in Medicare among patients with allergic conditions (HR = 1.69, 95% CI: 1.04, 2.75). While both exendin-based and human-analog GLP-1 RAs generated overlapping 95% CIs, the use of the former was associated with a moderately elevated HR (HR = 1.38, 95% CI: 0.94, 2.02); this finding was primarily driven by an increased risk in the CPRD (HR = 2.19, 95% CI: 1.10, 4.36). Stratifying by duration of follow-up resulted in variable HRs with wide 95% CIs (<30 days: HR = 0.86 (95% CI: 0.41, 1.83); 30–365 days: HR = 1.19 (95% CI: 0.90, 1.58); >365 days: HR = 1.10 (95% CI: 0.53, 2.27)). Restricting the analyses to anaphylactic events occurring after at least 1 year of follow-up generated results similar to those of the primary analysis (HR = 1.12, 95% CI: 0.62, 2.04), although an increased risk was observed in the Optum database (HR = 1.98, 95% CI: 1.03, 3.83). Finally, the sensitivity analyses generated findings consistent with those of the primary analysis (Web Figure 5), with the exception of the analysis censoring on initiation of nonsteroidal antiinflammatory drugs and antibiotics during follow-up; this generated an HR below the null value with a wide 95% CI.
Table 2.
Hazard Ratios for the Association Between Use of Glucagon-Like Peptide 1 Receptor Agonists and Risk of an Anaphylactic Reaction When Compared With Dipeptidyl Peptidase 4 Inhibitors in 4 UK and US Databases, 2007–2019
| Glucagon-Like Peptide Receptor Agonists | Dipeptidyl Peptidase 4 Inhibitors | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Data Source |
No. of
Patients |
No. of
Events |
PY of
Follow-up |
IR a |
No. of
Patients |
No. of
Events |
PY of
Follow-up |
IR a |
Weighted
HR |
95% CI |
| CPRD | 11,502 | 13 | 26,417 | 49.2 | 111,153 | 76 | 244,235 | 30.9 | 1.61 | 0.89, 2.90 |
| Medicare | 74,442 | 30 | 74,267 | 40.4 | 745,457 | 350 | 1,074,684 | 32.6 | 1.20 | 0.82, 1.77 |
| Optumb | 90,966 | 33 | 81,747 | 40.4 | 274,922 | 105 | 312,545 | 33.6 | 1.19 | 0.76, 1.86 |
| IBM MarketScanc | 147,188 | 46 | 147,745 | 31.1 | 509,988 | 198 | 635,755 | 31.1 | 0.97 | 0.69, 1.38 |
| Pooled cohortsd | 324,098 | 122 | 330,177 | 36.9 | 1,641,520 | 729 | 2,267,219 | 32.1 | 1.15 | 0.94, 1.42 |
Abbreviations: CI, confidence interval; CPRD, Clinical Practice Research Datalink; HR, hazard ratio; IR, incidence rate; PY, person-years.
a Number of cases per 100,000 PY.
b Optum, Inc., Eden Prairie, Minnesota.
c IBM, Armonk, New York.
d I 2 = 0%.
The secondary comparison included 259,929 and 366,067 patients newly treated with GLP-1 RAs and SGLT-2 inhibitors, respectively. The CPRD did not contribute to this analysis because of low event counts. Before propensity score fine stratification weighting, the exposure groups were generally similar across most characteristics, although there were more females among GLP-1 RA users than among SGLT-2 inhibitor users (Table 3). After propensity score fine stratification weighting, the baseline characteristics were well-balanced in the individual databases and after pooling (Table 3 and Web Table 4). The cumulative incidence curves of the exposure groups diverged in the Optum and MarketScan databases (Web Figure 6). Overall, the use of GLP-1 RAs was associated with a 38% increased risk of anaphylactic reaction, compared with the use of SGLT-2 inhibitors (40.7/100,000 person-years vs. 29.4/100,000 person-years, respectively; HR = 1.38, 95% CI: 1.02, 1.87) (Table 4). Stratification by follow-up duration did not reveal increased risk in any particular time period (<30 days: HR = 2.11 (95% CI: 0.61, 7.30); 30–365 days: HR = 1.21 (95% CI: 0.83, 1.78); >365 days: HR = 1.37 (95% CI: 0.77, 2.44)).
Table 3.
Baseline Characteristics of Initiators of Glucagon-Like-Peptide 1 Receptor Agonists and Sodium-Glucose Cotransporter 2 Inhibitors Before and After Propensity Score Fine Stratification Weighting in 3 US Databases, 2013–2019
| Before Weighting | After Weighting | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Characteristic |
GLP-1 RAs
(n = 262,462) |
SGLT-2 Inhibitors
a
(n = 368,963) |
ASD |
GLP-1 RAs
(n = 259,929) |
SGLT-2 Inhibitors
a
(n = 366,067) |
ASD | ||||
| No. | % | No. | % | No. | % | No. | % | |||
| Age, yearsb | 60.8 (9.8) | 61.7 (9.5) | 0.09 | 60.9 (9.8) | 60.9 (9.7) | 0.00 | ||||
| Female sex | 141,948 | 54.1 | 163,211 | 44.2 | 0.20 | 140,789 | 54.2 | 198,846 | 54.3 | 0.00 |
| Ever smoking | 36,619 | 13.9 | 44,746 | 12.1 | 0.05 | 36,450 | 14.0 | 48,574 | 13.3 | 0.02 |
| Alcohol dependence | 2,243 | 0.8 | 3,324 | 0.9 | 0.00 | 2,232 | 0.9 | 3,025 | 0.8 | 0.00 |
| Drug dependence | 4,229 | 1.6 | 4,353 | 1.9 | 0.04 | 4,206 | 1.6 | 5,622 | 1.5 | 0.01 |
| Obesity | 92,786 | 35.3 | 96,393 | 26.1 | 0.20 | 92,079 | 35.4 | 127,982 | 35.0 | 0.01 |
| Cancer | 28,086 | 10.7 | 37,229 | 10.1 | 0.02 | 27,963 | 10.8 | 37,843 | 10.3 | 0.01 |
| Chemotherapy | 8,613 | 3.3 | 10,924 | 3.0 | 0.02 | 8,588 | 3.3 | 11,802 | 3.2 | 0.00 |
| Chronic kidney disease | 325 | 0.1 | 401 | 0.1 | 0.00 | 324 | 0.1 | 450 | 0.1 | 0.00 |
| Anaphylactic reaction | 33,626 | 12.8 | 26,715 | 7.2 | 0.19 | 33,474 | 12.9 | 43,445 | 11.9 | 0.03 |
| Angioedema | 783 | 0.3 | 990 | 0.3 | 0.01 | 782 | 0.3 | 1,147 | 0.3 | 0.00 |
| Stevens-Johnson syndrome and TEN | 20 | 0.0 | 15 | 0.0 | 0.00 | 20 | 0.0 | 20 | 0.0 | 0.00 |
| Mastocytosis | 16 | 0.0 | 12 | 0.0 | 0.00 | 16 | 0.0 | 14 | 0.0 | 0.00 |
| Asthma | 24,664 | 9.4 | 26,102 | 7.1 | 0.08 | 24,563 | 9.5 | 33,858 | 9.2 | 0.01 |
| Drug allergies | 13,652 | 5.2 | 14,827 | 4.0 | 0.06 | 13,604 | 5.2 | 17,790 | 4.9 | 0.02 |
| Eczema | 11,955 | 4.5 | 16,137 | 4.4 | 0.01 | 11,907 | 4.6 | 16,627 | 4.5 | 0.00 |
| Food allergy | 2,662 | 1.0 | 3,282 | 0.9 | 0.01 | 2,654 | 1.0 | 3,629 | 1.0 | 0.00 |
| Allergic rhinitis | 2,718 | 1.0 | 3,806 | 1.0 | 0.00 | 2,707 | 1.0 | 3,712 | 1.0 | 0.00 |
| Allergic conjunctivitis | 28,684 | 10.9 | 37,157 | 10.1 | 0.03 | 28,563 | 11.0 | 39,999 | 10.9 | 0.00 |
| Urticaria | 2,434 | 0.9 | 3,098 | 0.8 | 0.01 | 2,420 | 0.9 | 3,394 | 0.9 | 0.00 |
| Hymenoptera venom allergy | 209 | 0.1 | 192 | 0.0 | 0.01 | 208 | 0.1 | 234 | 0.1 | 0.01 |
| Antihistamines | ||||||||||
| Nonoral | 13,990 | 5.3 | 16,310 | 4.4 | 0.04 | 13,950 | 5.4 | 19,698 | 5.4 | 0.00 |
| Oral | 16,045 | 6.1 | 19,862 | 5.4 | 0.03 | 16,041 | 6.2 | 23,089 | 6.3 | 0.01 |
| Corticosteroids | ||||||||||
| Oral | 49,743 | 18.9 | 60,154 | 16.3 | 0.07 | 49,734 | 19.1 | 70,272 | 19.2 | 0.00 |
| Orally inhaled | 18,178 | 6.9 | 20,151 | 5.5 | 0.06 | 18,172 | 7.0 | 25,218 | 6.9 | 0.00 |
| Epinephrine injection | 3,857 | 1.5 | 4,669 | 1.3 | 0.02 | 3,854 | 1.5 | 5,402 | 1.5 | 0.00 |
| Biologics | 9,677 | 3.7 | 11,334 | 3.1 | 0.03 | 9,638 | 3.7 | 13,136 | 3.6 | 0.01 |
| Stable angina | 10,152 | 3.9 | 14,012 | 3.8 | 0.00 | 10,115 | 3.9 | 13,597 | 3.7 | 0.01 |
| Unstable angina | 4,901 | 1.9 | 6,561 | 1.8 | 0.01 | 4,878 | 1.9 | 6,676 | 1.8 | 0.00 |
| Myocardial infarction | 9,495 | 3.6 | 12,674 | 3.4 | 0.01 | 9,452 | 3.6 | 12,651 | 3.5 | 0.01 |
| Other coronary atherosclerotic diseases | 48,477 | 18.5 | 64,858 | 17.6 | 0.02 | 48,285 | 18.6 | 65,188 | 17.8 | 0.02 |
| Heart failure | 19,470 | 7.4 | 20,753 | 5.6 | 0.07 | 19,406 | 7.5 | 25,793 | 7.0 | 0.02 |
| Previous cardiac procedure | 1,991 | 0.8 | 3,208 | 0.9 | 0.01 | 1,980 | 0.8 | 2,751 | 0.7 | 0.00 |
| Peripheral circulatory disorders | 15,666 | 6.0 | 20,078 | 5.4 | 0.02 | 32,301 | 12.4 | 43,104 | 11.8 | 0.02 |
| Stroke | 17,290 | 6.6 | 22,780 | 6.2 | 0.02 | 17,218 | 6.6 | 23,073 | 6.3 | 0.01 |
| Transient ischemic attack | 4,445 | 1.7 | 5,679 | 1.5 | 0.01 | 4,431 | 1.7 | 6,011 | 1.6 | 0.00 |
| Neuropathy | 48,624 | 18.5 | 57,200 | 15.5 | 0.08 | 48,301 | 18.6 | 64,986 | 17.7 | 0.02 |
| Gastroparesis | 1,288 | 0.5 | 1,635 | 0.4 | 0.01 | 1,281 | 0.5 | 2,044 | 0.6 | 0.01 |
| Retinopathy | 15,886 | 6.0 | 20,919 | 5.7 | 0.02 | 15,738 | 6.0 | 20,853 | 5.7 | 0.02 |
| Nephropathy | 28,671 | 10.9 | 27,724 | 7.5 | 0.12 | 28,494 | 11.0 | 36,407 | 9.9 | 0.03 |
| No. of antidiabetic drugs at cohort entrya | 0.9 (0.7) | 1 (0.7) | 0.10 | 0.9 (0.7) | 0.9 (0.7) | 0.00 | ||||
| Metformin | ||||||||||
| Current use | 91,865 | 35.0 | 136,523 | 37.0 | 0.04 | 91,858 | 35.3 | 131,098 | 35.8 | 0.01 |
| Past use | 193,063 | 73.6 | 288,659 | 78.2 | 0.11 | 193,049 | 74.3 | 273,731 | 74.8 | 0.01 |
| Sulfonylureas | ||||||||||
| Current use | 51,224 | 19.5 | 72,328 | 19.6 | 0.00 | 51,216 | 19.7 | 72,135 | 19.7 | 0.00 |
| Past use | 113,502 | 43.2 | 159,235 | 43.2 | 0.00 | 113,489 | 43.7 | 159,327 | 43.5 | 0.00 |
| Thiazolidinediones | ||||||||||
| Current use | 10,717 | 4.1 | 14,282 | 3.9 | 0.01 | 10,715 | 4.1 | 14,949 | 4.1 | 0.00 |
| Past use | 26,071 | 9.9 | 35,041 | 9.5 | 0.01 | 26,069 | 10.0 | 36,569 | 10.0 | 0.00 |
| DPP-4 inhibitors | ||||||||||
| Current use | 32,528 | 12.4 | 64,899 | 17.6 | 0.15 | 32,524 | 12.5 | 46,236 | 12.6 | 0.00 |
| Past use | 80,720 | 30.7 | 139,646 | 37.8 | 0.15 | 80,712 | 31.0 | 114,798 | 31.4 | 0.01 |
| α-Glucosidase inhibitors (ever use) | 1,664 | 0.6 | 2,434 | 0.6 | 0.00 | 1,663 | 0.6 | 2,283 | 0.6 | 0.00 |
| Meglitinides (ever use) | 4,428 | 1.7 | 6,233 | 1.7 | 0.00 | 4,427 | 1.7 | 6,131 | 1.7 | 0.00 |
| Obstructive sleep apnea | 48,680 | 18.5 | 47,406 | 12.8 | 0.16 | 48,390 | 18.6 | 68,067 | 18.6 | 0.00 |
| Osteoarthritis | 60,783 | 23.2 | 70,198 | 19.0 | 0.10 | 60,525 | 23.3 | 82,267 | 22.5 | 0.02 |
| Other arthropathies | 131,035 | 49.9 | 161,264 | 43.7 | 0.12 | 130,298 | 50.1 | 180,975 | 49.4 | 0.01 |
| Dorsopathies | 86,369 | 32.9 | 104,721 | 28.4 | 0.10 | 85,895 | 33.0 | 118,958 | 32.5 | 0.01 |
| Pancreatitis | 1,132 | 0.4 | 2,924 | 0.8 | 0.05 | 1,124 | 0.4 | 1,689 | 0.5 | 0.00 |
| Urinary tract infections | 43,173 | 16.4 | 47,357 | 12.8 | 0.10 | 38,064 | 14.6 | 51,429 | 14.0 | 0.02 |
| Delirium or psychosis | 4,993 | 1.9 | 5,711 | 1.5 | 0.03 | 4,978 | 1.9 | 6,741 | 1.8 | 0.01 |
| Dementia | 7,118 | 2.7 | 9,018 | 2.4 | 0.02 | 7,086 | 2.7 | 9,455 | 2.6 | 0.01 |
| Angiotensin-converting enzyme inhibitors | 112,345 | 42.8 | 160,362 | 43.5 | 0.01 | 112,337 | 43.2 | 158,371 | 43.3 | 0.00 |
| Angiotensin II receptor blockers | 80,489 | 30.7 | 111,814 | 30.3 | 0.01 | 80,478 | 31.0 | 112,640 | 30.8 | 0.00 |
| Beta blockers | 91,442 | 34.8 | 121,033 | 32.8 | 0.04 | 91,432 | 35.2 | 126,192 | 34.5 | 0.01 |
| Calcium channel blockers | 67,453 | 25.7 | 92,833 | 25.2 | 0.01 | 67,444 | 25.9 | 93,432 | 25.5 | 0.01 |
| Diuretics | 77,386 | 29.5 | 83,355 | 22.6 | 0.16 | 77,376 | 29.8 | 106,811 | 29.2 | 0.01 |
| Other antihypertensive drugs | 15,852 | 6.0 | 17,632 | 4.8 | 0.06 | 15,848 | 6.1 | 21,582 | 5.9 | 0.01 |
| Antiplatelet agents | 24,008 | 9.1 | 35,268 | 9.6 | 0.01 | 24,005 | 9.2 | 33,647 | 9.2 | 0.00 |
| Statins | 172,361 | 65.7 | 247,841 | 67.2 | 0.03 | 172,348 | 66.3 | 240,373 | 65.7 | 0.01 |
| Warfarin | 8,814 | 3.4 | 9,136 | 2.5 | 0.05 | 8,811 | 3.4 | 11,855 | 3.2 | 0.01 |
| Oral anticoagulants (except warfarin) | 8,709 | 3.3 | 11,442 | 3.1 | 0.01 | 8,709 | 3.3 | 11,852 | 3.2 | 0.01 |
| Injectable anticoagulants | 1,840 | 0.7 | 1,758 | 0.5 | 0.03 | 1,840 | 0.7 | 2,509 | 0.7 | 0.00 |
| Nonsteroidal antiinflammatory drugs | 69,978 | 26.7 | 93,050 | 25.2 | 0.03 | 69,972 | 26.9 | 99,490 | 27.2 | 0.01 |
| Antibiotic use | ||||||||||
| Past use | 149,345 | 56.9 | 194,073 | 52.6 | 0.09 | 149,329 | 57.4 | 210,793 | 57.6 | 0.00 |
| Recent use | 30,237 | 11.5 | 38,302 | 10.4 | 0.04 | 30,232 | 11.6 | 42,683 | 11.7 | 0.00 |
| Nonbiological DMARDs | 7,348 | 2.8 | 9,097 | 2.5 | 0.02 | 7,348 | 2.8 | 10,109 | 2.8 | 0.00 |
| Bone mineral density | 16,063 | 6.1 | 18,695 | 5.1 | 0.05 | 16,001 | 6.2 | 21,461 | 5.9 | 0.01 |
| Colonoscopy | 26,097 | 9.9 | 34,829 | 9.4 | 0.02 | 25,983 | 10.0 | 36,470 | 10.0 | 0.00 |
| Influenza vaccine | 103,410 | 39.4 | 133,242 | 36.1 | 0.07 | 102,873 | 39.6 | 139,497 | 38.1 | 0.03 |
| Mammogram | 66,605 | 25.4 | 76,266 | 20.7 | 0.11 | 66,234 | 25.5 | 92,944 | 25.4 | 0.00 |
| Papanicolaou smear | 27,304 | 10.4 | 32,606 | 8.8 | 0.05 | 27,162 | 10.4 | 40,306 | 11.0 | 0.02 |
| Pneumococcal vaccine | 77,397 | 29.5 | 101,728 | 27.6 | 0.04 | 76,878 | 29.6 | 102,847 | 28.1 | 0.03 |
| Prostate examination (DRE) | 22,522 | 8.6 | 38,209 | 10.0 | 0.06 | 22,378 | 8.6 | 30,260 | 8.3 | 0.01 |
| Hospitalization | ||||||||||
| Recent (≤30 days prior to cohort entry) | 27,193 | 10.4 | 31,315 | 8.5 | 0.06 | 27,080 | 10.4 | 37,313 | 10.2 | 0.01 |
| Older (31–365 days prior to cohort entry) | 28,101 | 10.7 | 31,304 | 8.5 | 0.08 | 27,945 | 10.7 | 38,224 | 10.4 | 0.01 |
| Basic or comprehensive metabolic blood chemistry test | 218,835 | 83.4 | 304,276 | 82.5 | 0.02 | 217,317 | 83.6 | 301,757 | 82.4 | 0.03 |
| Frailty scoreb | 0.2 (0.1) | 0.1 (0.1) | 0.54 | 0.2 (0.1) | 0.2 (0.1) | 0.00 | ||||
| No. of cardiologist visitsb | 2.1 (5.9) | 2 (5.5) | 0.03 | 2.1 (5.9) | 2.2 (5.7) | 0.01 | ||||
| No. of distinct medication prescriptionsb | 11.7 (6.2) | 10.6 (5.8) | 0.18 | 11.8 (6.1) | 11.8 (6.4) | 0.00 | ||||
| No. of emergency department visitsb | 0.6 (1.9) | 0.5 (1.7) | 0.08 | 0.6 (1.9) | 0.6 (1.9) | 0.00 | ||||
| No. of endocrinologist visitsb | 1.2 (5.5) | 1 (5.0) | 0.04 | 1.2 (5.5) | 1.3 (6.3) | 0.01 | ||||
| No. of hemoglobinA1c tests orderedb | 2.3 (1.3) | 2.3 (1.3) | 0.00 | 2.3 (1.3) | 2.3 (1.3) | 0.00 | ||||
| No. of internal medicine/family medicine visitsb | 13.8 (18.2) | 13.1 (17.4) | 0.03 | 13.8 (18.3) | 13.9 (18.0) | 0.00 | ||||
| SES proxy measure: pharmacy copay (US dollars)b | 22.5 (27.3) | 22.7 (28.7) | 0.01 | 22.6 (26.6) | 22.6 (31.3) | 0.00 | ||||
Abbreviations: ASD, absolute standardized difference; DMARD, disease-modifying antirheumatic drug; DPP-4, dipeptidyl peptidase 4; DRE, digital rectal examination; GLP-1 RA, glucagon-like peptide receptor agonist; SD, standard deviation; SES, socioeconomic status; SGLT-2, sodium glucose cotransporter 2; TEN, toxic epidermal necrolysis.
a Pooling was conducted by taking the average of the mean values and proportions of the exposed group and taking a weighted average of the mean values and proportions of the unexposed group, the weight being the ratio of the number of exposed individuals in that database to the number of all exposed individuals.
b Values are expressed as mean (standard deviation).
Table 4.
Hazard Ratios for the Association Between Use of Glucagon-Like Peptide 1 Receptor Agonists and Risk of an Anaphylactic Reaction When Compared With Sodium-Glucose Cotransporter 2 Inhibitors in 4 UK and US Databases, 2013–2019
| Data Source | Glucagon-Like PeptideReceptor Agonists | Sodium-Glucose Cotransporter2 Inhibitors | Weighted HR | 95% CI | ||||||
|---|---|---|---|---|---|---|---|---|---|---|
|
No. of
Patients |
No. of
Events |
PY of
Follow-up |
IR a |
No. of
Patients |
No. of
Events |
PY of
Follow-up |
IR a | |||
| Medicare | 79,768 | 26 | 70,962 | 36.6 | 114,161 | 27 | 101,988 | 26.5 | 1.36 | 0.81, 2.42 |
| Optumb | 78,300 | 27 | 63,678 | 42.4 | 100,563 | 31 | 90,371 | 34.3 | 1.22 | 0.67, 2.22 |
| IBM MarketScanc | 101,861 | 39 | 90,970 | 42.9 | 151,343 | 44 | 154,813 | 28.4 | 1.48 | 0.94, 2.32 |
| Pooled cohortsd | 259,929 | 92 | 225,610 | 40.7 | 366,067 | 102 | 347,172 | 29.4 | 1.38 | 1.02, 1.87 |
Abbreviations: CI, confidence interval; HR, hazard ratio; IR, incidence rate; PY, person-years.
a Number of cases per 100,000 person-years.
b Optum, Inc., Eden Prairie, Minnesota.
c IBM, Armonk, New York.
d I 2 = 0%.
DISCUSSION
The results of this large, multisite population-based cohort study indicate that when compared with DPP-4 inhibitors, GLP-1 RAs were associated with a 15% numerical increase in the risk of an anaphylactic reaction, with 95% CIs including the null value. In contrast, the use of GLP-1 RAs was associated with a 38% increased risk when compared with SGLT-2 inhibitors.
Regulatory concerns regarding a possible association between GLP-1 RAs and anaphylactic reaction emerged following events reported in randomized clinical trials of GLP-1 RAs; in particular, pooled data from trials of taspoglutide, a human-analog GLP-1 RA, revealed an event rate of 4.3%, leading to discontinuation of the drug’s development (29). In contrast, the event rates were variable in other GLP-1 RA trials, ranging from 0% for exenatide to 0.2% for lixisenatide (29). However, the association has continued to be scrutinized during regulatory decisions on GLP-1 RAs (29). It is important to note that these trials were neither designed for nor sufficiently powered to assess rare events such as anaphylactic reaction, with even the largest of the trials accruing a sample size of only 14,752 patients (30). Moreover, the event adjudication process was inconsistent across the clinical development programs of various second- to third-line antidiabetic drugs, making it difficult to infer the comparative safety of GLP-1 RAs regarding anaphylactic reaction. Consequently, the frequency of the event in the natural setting of clinical practice, its timing with respect to drug initiation, and the strength of the association, if any, have remained unknown. Given that GLP-1 RAs are preferentially prescribed to patients with cardiovascular diseases and obesity (2)—conditions associated with higher fatality rates for anaphylactic reactions (31)—assessing this possible association merited a well-powered investigation.
To our knowledge, this is the first cohort study to have assessed the comparative safety of GLP-1 RAs with respect to anaphylactic reaction in the real-world setting. GLP-1 RAs are biologics with sequence differences from endogenous glucagon-like peptide 1 (ranging between 3% and 50%) that may precipitate anaphylactic reactions (5). Interestingly, the magnitude of the HR was higher when comparing GLP-1 RAs with SGLT-2 inhibitors than when comparing GLP-1 RAs with DPP-4 inhibitors, although the 95% CIs overlapped. From a biological standpoint, these different effect sizes may relate to the pharmacological profiles of these drug classes. The inhibition of the immunologically active DPP-4 enzyme by DPP-4 inhibitors may up-regulate allergenic pathways, although the mechanism of precipitating anaphylactic reactions is unclear (32). In contrast, SGLT-2 inhibitors have not been associated with proallergenic mechanisms. Importantly, unlike DPP-4 inhibitors and SGLT-2 inhibitors, GLP-1 RAs are administered parenterally, a route that is known to precipitate drug allergies more frequently than oral administration (33). It is unknown whether the current association also applies to oral GLP-1 RAs, and thus future studies will be needed to examine this association when the use of these formulations becomes more widespread. Finally, we did not examine the biological reasons behind the effects of GLP-1 RAs—whether they are due to GLP-1 RA itself or to some other pharmacological or other mediator. In future research, investigators should distinguish between direct and indirect effects of GLP-1 RA on this association.
One unanswered regulatory question has been whether anaphylactic reactions precipitated by GLP-1 RAs are a class or molecule-specific effect (6). Indeed, the structural similarity of the different GLP-1 RAs to human glucagon-like peptide 1 varies, with exendin-based GLP-1 RAs having 50%–53% homology and human-analog GLP-1 RAs having at least 90% homology (4). These structural differences translate into a higher prevalence of antidrug antibodies with exendin-based GLP-1 RAs (ranging from 44% to 70%) than with human-analog GLP-1 RAs (less than 10%) (3). An analysis using the World Health Organization’s VigiBase database found that the reporting odds of anaphylactic reactions with exendin-based GLP-1 RAs were double those of human analog GLP-1 RAs (reporting odds ratio = 2.08, 95% CI: 1.37, 3.19) (13). In the present study, neither exendin-based nor human-analog GLP-1 RAs were associated with a statistically significantly increased risk of anaphylactic reaction. However, exendin-based GLP-1 RAs generated a more elevated HR, a finding driven by a statistically significant association in the CPRD. We also observed that the cumulative incidence curves diverged months to years after the first GLP-1 RA dose, corroborating the relatively late onset of the event described in one case report (10 months after the first dose of GLP-1 RA) (9). This delayed effect might be explained by several factors. First, this may be due to the reintroduction of GLP-1 RAs after a short treatment gap (i.e., within 90 days as per the grace period used between nonoverlapping consecutive prescriptions). Second, it is possible that the observed effects are due to the persistence of antidrug antibodies after long-term administration of the drug (20, 34). Finally, it is also possible that exposure to GLP-1 RAs primes the body to an anaphylactogenic effect of other mediators introduced after cohort entry. Importantly, we did not find an overall increased risk in subgroup analyses based on age and history of allergic conditions, although an increased risk was observed with the latter in Medicare. Additional studies will be needed to investigate the interaction between old age and allergic conditions in the association between GLP-1 RA and anaphylactic reactions.
Our study had several strengths. First, the use of 4 population-based databases from 2 countries maximized the generalizability of our findings across different health systems, age groups, and socioeconomic strata. Second, we compared GLP-1 RAs with 2 important therapeutic alternatives, thereby increasing the clinical relevance of our findings. Finally, the results of our primary analysis were highly consistent across multiple sensitivity analyses.
Our study also had some limitations. First, the outcome definition for anaphylactic reaction may have been subject to some misclassification. However, this misclassification would likely have biased the effect estimates towards the null, and thus our estimates should be interpreted as conservative. Second, there may have been exposure misclassification. However, in the United Kingdom, general practitioners are responsible for the long-term care of patients with diabetes, while Optum, MarketScan, and Medicare record prescription data from both general practitioners and specialists. As such, exposure misclassification is likely to have been minimal and nondifferential between the exposure groups. Third, because the primary analysis censored follow-up at the time of treatment discontinuation, informative censoring is possible. However, anaphylactic reactions are not usually preceded by milder forms of allergic reactions, which may result in the discontinuation of a drug that has been linked with the outcome of interest. Reassuringly, we observed similar results when using an intention-to-treat exposure definition in a sensitivity analysis. Fourth, our analysis censoring on use of nonsteroidal antiinflammatory drugs and antibiotics, 2 common pharmacological triggers of anaphylactic reactions, yielded a point estimate below the null with a wide 95% CI (with an upper limit of 1.42). This may have been due to the truncation of follow-up, given the frequent initiation of these drugs in this population. Finally, as with any observational study, residual confounding is possible. However, our propensity score model included a wide range of variables, and the exposure groups were well balanced for all covariates after propensity score fine stratification weighting. Moreover, confounding by indication is unlikely to be an important issue, given that risk of anaphylactic reaction is not a clinical consideration when prescribing GLP-1 RAs versus other second- to third-line antidiabetic drugs.
In summary, the results of this large population-based study indicate that GLP-1 RAs may be associated with a modestly increased risk of anaphylactic reaction compared with DPP-4 inhibitors and SGLT-2 inhibitors. While these findings corroborate previous case reports (8–13), the rarity of this outcome and the modest associations observed in this study provide some reassurance and should be balanced with the known clinical benefits of this class of medications. Future studies should examine whether this association is brought about by pharmacological and nonpharmacological mediating factors.
Supplementary Material
ACKNOWLEDGMENTS
Author affiliations: Department of Epidemiology, Biostatistics, and Occupational Health, Faculty of Medicine and Health Sciences, McGill University, Montreal, Quebec, Canada (Richeek Pradhan, Christel Renoux, Laurent Azoulay); Center for Clinical Epidemiology, Lady Davis Institute, Jewish General Hospital, Montreal, Quebec, Canada (Richeek Pradhan, Hui Yin, Christina Santella, Christel Renoux, Laurent Azoulay); Division of Pharmacoepidemiology, Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School, Boston, Massachusetts, United States (Elisabetta Patorno, Helen Tesfaye, Sebastian Schneeweiss, Jessica Franklin, Ajinkya Pawar); Division of Endocrinology, Jewish General Hospital, Montreal, Quebec, Canada (Oriana H. Y. Yu); Department of Neurology and Neurosurgery, Faculty of Medicine and Health Sciences, McGill University, Montreal, Quebec, Canada (Christel Renoux); and Gerald Bronfman Department of Oncology, Faculty of Medicine and Health Sciences, McGill University, Montreal, Quebec, Canada (Laurent Azoulay).
This work was supported by a Foundation Scheme grant from the Canadian Institutes of Health Research (grant FDN-143328) and the Division of Pharmacoepidemiology and Pharmacoeconomics, Department of Medicine, Brigham and Women’s Hospital, Harvard Medical School (Boston, Massachusetts). R.P. receives doctoral funding from the Fonds de Recherche du Québec-Santé. E.P. is supported by a career development grant (grant K08AG055670) from the National Institute on Aging, US National Institutes of Health. The work of J.F. is funded by National Institutes of Health grant R01HL141505. O.H.Y.Y. holds a Chercheur-Boursier Clinicien Junior 1 award from the Fonds de Recherche du Québec-Santé. C.R. is the recipient of a Chercheur Boursier Junior 2 award from the Fonds de Recherche du Québec-Santé. L.A. holds a Chercheur-Boursier Senior award from the Fonds de Recherche du Québec-Santé and is the recipient of a William Dawson Scholar Award from McGill University.
No additional data are available.
Parts of this study were presented at the 36th Annual International Conference on Pharmacoepidemiology & Therapeutic Risk Management (virtual), September 16–17, 2020.
The sponsors had no influence on the design and conduct of the study; the collection, management, analysis, and interpretation of the data; or the preparation, review, or approval of the manuscript.
E.P. is a co–principal investigator of an investigator-initiated grant to the Brigham and Women’s Hospital from Boehringer Ingelheim (Ingelheim am Rhein, Germany), unrelated to the topic of this work. S.S. is a co–principal investigator of investigator-initiated grants to the Brigham and Women’s Hospital from Union Chimique Belge (Brussels, Belgium) and Boehringer Ingelheim, unrelated to the topic of this study. He is a consultant to Aetion, Inc. (New York, New York), a software manufacturer in which he owns equity. His interests were declared, reviewed, and approved by the Brigham and Women’s Hospital and Partners HealthCare System in accordance with their institutional compliance policies. L.A. received consulting fees from Janssen Pharmaceuticals (Raritan, New Jersey) and Pfizer, Inc. (New York, New York) unrelated to this project. The other authors have no potential conflicts of interest to declare.
REFERENCES
- 1. Bethel MA, Patel RA, Merrill P, et al. Cardiovascular outcomes with glucagon-like peptide-1 receptor agonists in patients with type 2 diabetes: a meta-analysis. Lancet Diabetes Endocrinol. 2018;6(2):105–113. [DOI] [PubMed] [Google Scholar]
- 2. American Diabetes Association . 8. Pharmacologic approaches to glycemic treatment: Standards of Medical Care in Diabetes—2018. Diabetes Care. 2018;41(suppl 1):S73–S85. [DOI] [PubMed] [Google Scholar]
- 3. Filippatos TD, Panagiotopoulou TV, Elisaf MS. Adverse effects of GLP-1 receptor agonists. Rev Diabetic Stud. 2014;11(3-4):202–230. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Tran KL, Park YI, Pandya S, et al. Overview of glucagon-like peptide-1 receptor agonists for the treatment of patients with type 2 diabetes. Am Health Drug Benefits. 2017;10(4):178–188. [PMC free article] [PubMed] [Google Scholar]
- 5. Lund A, Knop FK, Vilsbøll T. Glucagon-like peptide-1 receptor agonists for the treatment of type 2 diabetes: differences and similarities. Eur J Intern Med. 2014;25(5):407–414. [DOI] [PubMed] [Google Scholar]
- 6. Center for Drug Evaluation and Research, Food and Drug Administration . Application Number: 208471Orig1s000. Risk Assessment and Risk Mitigation Review(s). (Risk assessment and risk mitigation review for lixisenatide). Beltsville, MD: US Food and Drug Administration; 2016. https://www.accessdata.fda.gov/drugsatfda_docs/nda/2016/208471Orig1s000RiskR.pdf. Accessed November 1, 2020. [Google Scholar]
- 7. Rosenstock J, Balas B, Charbonnel B, et al. The fate of taspoglutide, a weekly GLP-1 receptor agonist, versus twice-daily exenatide for type 2 diabetes: the T-Emerge 2 Trial. Diabetes Care. 2013;36(3):498–504. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Shah M, Kanunga J. Anaphylactic reaction due to exenatide: a case report. Ann Allergy Asthma Immunol. 2008;100(suppl 1):38. [Google Scholar]
- 9. Ornelas C, Caiado J, Lopes A, et al. Anaphylaxis to long-acting release exenatide. J Investig Allergol Clin Immunol. 2018;28(5):332–334. [DOI] [PubMed] [Google Scholar]
- 10. Pérez E, Martínez-Tadeo J, Callero A, et al. A case report of allergy to exenatide. J Allergy Clin Immunol Pract. 2014;2(6):822–823. [DOI] [PubMed] [Google Scholar]
- 11. Steveling EH, Winzeler B, Bircher AJ. Systemic allergic reaction to the GLP-1 receptor agonist exenatide. J Pharm Technol. 2014;30(5):182–186. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Shamriz O, Nasereddin A, Mosenzon O, et al. Allergic reaction to exenatide and lixisenatide but not to liraglutide: unveiling anaphylaxis to glucagon-like peptide 1 receptor agonists. Diabetes Care. 2019;42(9):e141–e142. [DOI] [PubMed] [Google Scholar]
- 13. Pradhan R, Montastruc F, Rousseau V, et al. Exendin-based glucagon-like peptide-1 receptor agonists and anaphylactic reactions: a pharmacovigilance analysis. Lancet Diabetes Endocrinol. 2020;8(1):13–14. [DOI] [PubMed] [Google Scholar]
- 14. Karagiannis T, Boura P, Tsapas A. Safety of dipeptidyl peptidase 4 inhibitors: a perspective review. Ther Adv Drug Saf. 2014;5(3):138–146. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. McGill JB, Subramanian S. Safety of sodium-glucose co-transporter 2 inhibitors. Am J Cardiol. 2019;124(suppl 1):S45–S52. [DOI] [PubMed] [Google Scholar]
- 16. Curtis HJ, Dennis JM, Shields BM, et al. Time trends and geographical variation in prescribing of drugs for diabetes in England from 1998 to 2017. Diabetes Obes Metab. 2018;20(9):2159–2168. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Turner PJ, Jerschow E, Umasunthar T, et al. Fatal anaphylaxis: mortality rate and risk factors. J Allergy Clin Immunol Pract. 2017;5(5):1169–1178. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Desai S, Brinker A, Swann J, et al. Sitagliptin-associated drug allergy: review of spontaneous adverse event reports. Arch Intern Med. 2010;170(13):1169–1171. [DOI] [PubMed] [Google Scholar]
- 19. Heinzerling L, Raile K, Rochlitz H, et al. Insulin allergy: clinical manifestations and management strategies. Allergy. 2008;63(2):148–155. [DOI] [PubMed] [Google Scholar]
- 20. Fineman MS, Mace KF, Diamant M, et al. Clinical relevance of anti-exenatide antibodies: safety, efficacy and cross-reactivity with long-term treatment. Diabetes Obes Metab. 2012;14(6):546–554. [DOI] [PubMed] [Google Scholar]
- 21. Peng MM, Jick H. A population-based study of the incidence, cause, and severity of anaphylaxis in the United Kingdom. Arch Intern Med. 2004;164(3):317–319. [DOI] [PubMed] [Google Scholar]
- 22. Walsh KE, Cutrona SL, Foy S, et al. Validation of anaphylaxis in the Food and Drug Administration’s Mini-Sentinel. Pharmacoepidemiol Drug Saf. 2013;22(11):1205–1213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Caspard H, Steffey A, Mallory RM, et al. Evaluation of the safety of live attenuated influenza vaccine (LAIV) in children and adolescents with asthma and high-risk conditions: a population-based prospective cohort study conducted in England with the Clinical Practice Research Datalink. BMJ Open. 2018;8(12):e023118. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Kawai AT, Li L, Kulldorff M, et al. Absence of associations between influenza vaccines and increased risks of seizures, Guillain-Barré syndrome, encephalitis, or anaphylaxis in the 2012–2013 season. Pharmacoepidemiol Drug Saf. 2014;23(5):548–553. [DOI] [PubMed] [Google Scholar]
- 25. Wang C, Graham DJ, Kane RC, et al. Comparative risk of anaphylactic reactions associated with intravenous iron products. JAMA. 2015;314(19):2062–2068. [DOI] [PubMed] [Google Scholar]
- 26. Desai RJ, Rothman KJ, Bateman BT, et al. A propensity-score-based fine stratification approach for confounding adjustment when exposure is infrequent. Epidemiology. 2017;28(2):249–257. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Kubo S, Nakayama T, Matsuoka K, et al. Long term maintenance of IgE-mediated memory in mast cells in the absence of detectable serum IgE. J Immunol. 2003;170(2):775–780. [DOI] [PubMed] [Google Scholar]
- 28. Montañez MI, Mayorga C, Bogas G, et al. Epidemiology, mechanisms, and diagnosis of drug-induced anaphylaxis. Front Immunol. 2017;8:614. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Center for Drug Evaluation and Research, Food and Drug Administration . Application Number: 125469Orig1s000. Medical Review(s). (Medical review for dulaglutide). Beltsville, MD: US Food and Drug Administration; 2014. https://www.accessdata.fda.gov/drugsatfda_docs/nda/2014/125469Orig1s000MedRedt.pdf. Accessed November 1, 2020. [Google Scholar]
- 30. Cefalu WT, Kaul S, Gerstein HC, et al. Cardiovascular outcomes trials in type 2 diabetes: where do we go from here? Reflections from a Diabetes Care editors’ expert forum. Diabetes Care. 2018;41(1):14–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Turner PJ, Gowland MH, Sharma V, et al. Increase in anaphylaxis-related hospitalizations but no increase in fatalities: an analysis of United Kingdom national anaphylaxis data, 1992–2012. J Allergy Clin Immunol. 2015;135(4):956–963.e1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Tasic T, Bäumer W, Schmiedl A, et al. Dipeptidyl peptidase IV (DPP4) deficiency increases Th1-driven allergic contact dermatitis. Clin Exp Allergy. 2011;41(8):1098–1107. [DOI] [PubMed] [Google Scholar]
- 33. Warrington R, Silviu-Dan F, Wong T. Drug allergy. Allergy Asthma Clin Immunol. 2018;14(suppl 2):60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Buse JB, Henry RR, Han J, et al. Effects of exenatide (exendin-4) on glycemic control over 30 weeks in sulfonylurea-treated patients with type 2 diabetes. Diabetes Care. 2004;27(11):2628–2635. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.