Pancreatitis Risk In Patients Taking Antipsychotic Medications
Overview
This extensive nationwide study, spanning from 2006 to 2019, investigated the potential link between antipsychotic drug use and acute pancreatitis risk. Analyzing data from 52,006 cases and 518,081 controls, the study initially suggested an increased risk associated with both first-generation and second-generation antipsychotic drugs, particularly in past users. However, after adjusting for factors like alcohol abuse and comorbidities, the association was substantially weakened. Notably, only past use of first-generation agents maintained a statistically significant association with acute pancreatitis. The findings indicate that the observed association in previous case reports may be influenced by confounding factors, emphasizing the importance of comprehensive studies in understanding such medical correlations.
Introduction
This study delves into the diverse clinical manifestations of acute pancreatitis, ranging from mild, self-limiting cases to severe or fatal conditions. While established risk factors include gallstone disease and alcohol abuse, the potential link between medical drugs, specifically antipsychotic medications, and acute pancreatitis remains underexplored. Existing literature, primarily composed of case reports, suggests an association, especially with second-generation agents like Clozapine and Olanzapine. However, sparse and inconsistent data from controlled pharmacoepidemiological studies contribute to the ambiguity. Notably, previous research presents conflicting results, with some studies identifying an increased risk with first-generation phenothiazines, while others find no association or implicate only specific antipsychotic subtypes. This nationwide case–control study, encompassing a significantly larger dataset than previous investigations, aims to comprehensively analyze the potential associations between both first-generation and second-generation antipsychotic drugs and the risk of acute pancreatitis. Additionally, the study explores potential variations in these associations concerning the type of acute pancreatitis, distinguishing between gallstone-related and non-gallstone-related cases.
Methods
This study is based on data from the Swedish Pancreatitis Cohort (SwePan), a nationwide cohort spanning 1990 to 2019, focusing on acute pancreatitis incidence, long-term complications, and risk factors. Examining 52,006 incident cases between 2006 and 2019, this case–control study utilizes several Swedish registers. The study identifies cases through the National Patient Register, with comprehensive national coverage since 1987, and control subjects are matched based on age, sex, and municipality of residence. Ethical approval was obtained, and the study period ensures complete coverage by the National Prescribed Drug Register since July 1, 2005. To minimize exposure misclassification, individuals with a history of pancreatitis or pancreatic cancer are excluded.
The exposure assessment involves data on dispensed prescriptions for antipsychotic drugs (ATC code N05A). The study categorizes first-generation and second-generation agents, further classifying first-generation agents based on potency. Notably, no case or control subject received Penfluridol. Antipsychotic drug use is classified as “current,” “past,” or “never,” with the 90-day cut-off reflecting Sweden’s prescription dispensing practices.
The covariate assessment includes matching variables, country of birth, educational level, Charlson comorbidity index, alcohol abuse, use of antidepressants and mood stabilizers, total number of dispensed drugs, and psychiatric condition. Data is sourced from various registers, and diseases or conditions are identified by recorded diagnoses in the National Patient Register before the index date. The study employs rigorous methodology, considering potential confounders, to comprehensively analyze the association between antipsychotic drug use and acute pancreatitis risk. Limitations include expected misclassification due to the register-based design and unavoidable residual confounding.
Statistical Analysis
The study employed conditional logistic regression to evaluate the odds ratios (ORs) and corresponding 95% confidence intervals (CIs) associated with acute pancreatitis in relation to the use of first-generation and second-generation antipsychotic drugs. First-generation agents underwent additional analysis based on their potency to block dopamine receptors, categorized as low, intermediate, and high. The second-generation agents Clozapine and Olanzapine were treated as a distinct exposure category.
Two statistical models were constructed: Model 1, including only matching variables (age, sex, and municipality of residence), and Model 2, which added country of birth, educational level, Charlson comorbidity index, alcohol abuse, and psychiatric condition. To address collinearity, analyses were restricted to case and control subjects without current use of mood stabilizers and antidepressants in a separate model. Another model aimed to minimize polypharmacy bias by considering individuals with a prescription history of five or fewer unique drugs in the 6 months preceding the index date. Both regression models mutually adjusted for first-generation and second-generation antipsychotic drugs.
Sensitivity analyses were conducted to explore potential exposure misclassification during the register’s initial year, limiting the study period to January 1, 2007, to December 31, 2019. Additional sensitivity analyses altered the definition of current use of antipsychotic drugs to within;
(i) 100 days
(ii) 90 days plus a margin of 14 days
(iii) 100 days plus a margin of 14 days.
Subgroup analyses were performed based on the type of acute pancreatitis (gallstone-related or non-gallstone-related) and psychiatric condition (none or psychosis and/or bipolar or unipolar depression) to assess confounding by indication. Furthermore, subgroup analyses were conducted by age (<64 or ≥64 years) and sex. The study implemented robust methodologies to explore the nuanced associations between antipsychotic drug use and acute pancreatitis, considering various potential confounders and utilizing sensitivity analyses to validate results.
Results
The analytical cohort, comprising 52,006 cases of acute pancreatitis and 518,081 control subjects, exhibited distinct demographic, comorbid, and pharmaceutical characteristics. Case subjects, in comparison to control subjects, had slightly lower educational attainment and a significantly higher prevalence of comorbid conditions, particularly alcohol abuse (9.0% vs. 2.9%). Expectedly, individuals with acute pancreatitis displayed higher overall drug use, including antipsychotics, mood stabilizers, and antidepressants, compared to control subjects.
The prescribed and dispensed antipsychotic drugs in the study population were detailed, with Levomepromazine, Haloperidol, Flupenthixole, Risperidone, Quetiapine, and Olanzapine being among the most common agents. Psychiatric diagnoses, particularly psychosis and/or bipolar or unipolar depression, were associated with a greater comorbidity burden, emphasizing the substantial health complexity in the main patient groups for antipsychotic drug treatment.
The odds ratios (ORs) of acute pancreatitis related to antipsychotic drug use were presented. In the crude model, accounting for matching variables, both first-generation and second-generation agents showed an increased risk, with slightly higher ORs for past use compared to current use. However, the multivariable model substantially attenuated the ORs, resulting in a statistically significant association remaining only for past use of first-generation agents (OR 1.18 [95% CI 1.10–1.26]). Additional categorization of past use based on time since the last dispensed prescription was provided.
Confounding adjustments for co-current use of mood stabilizers and antidepressants, as well as polypharmacy, had minimal impact on the results. Subgroup analysis revealed an increased risk of acute pancreatitis for first-generation agents with low potency to block dopamine receptors but not for those with intermediate or high potency. Additionally, examining Clozapine and Olanzapine as a separate exposure category indicated an elevated risk for past use but not for current use (multivariable-adjusted OR 1.26 [1.14–1.39] and 0.93 [0.81–1.07], respectively). These findings contribute nuanced insights into the association between antipsychotic drug use and acute pancreatitis, emphasizing the relevance of specific agents and usage patterns.
Conclusion
In this nationwide case–control study, an overall lack of association was observed between second-generation antipsychotic drugs and the risk of acute pancreatitis after adjusting for potential confounders. Conversely, first-generation antipsychotic drugs, particularly those with low potency in blocking dopamine receptors, demonstrated a slightly increased risk of acute pancreatitis overall and specifically in non-gallstone-related cases, even after accounting for potential confounders.
Despite the prevalence of case reports suggesting a link between second-generation antipsychotic drugs and acute pancreatitis, well-controlled pharmacoepidemiological studies have not supported this association. Previous studies by Gasse et al. in Denmark (3083 patients, none exposed to second-generation agents) and Bodén et al. in Sweden (6161 cases, 110 exposed to second-generation agents) found no association between second-generation antipsychotic drugs and the risk of acute pancreatitis. Despite these findings, some product information for several second-generation antipsychotic drugs lists acute pancreatitis as a potential adverse reaction.
Consistent with the aforementioned studies, the current investigation revealed a positive association in crude models and a null association in multivariable models for second-generation antipsychotic drugs concerning acute pancreatitis (52,006 cases, 1787 exposed to second-generation agents). This pattern was consistent across subtypes of acute pancreatitis, including gallstone-related and non-gallstone-related cases. Notably, current users of Clozapine and Olanzapine, the second-generation agents most frequently implicated in case reports, did not exhibit an increased risk of acute pancreatitis. In summary, while adverse reactions cannot be categorically ruled out, especially given reports of disease recurrence after drug rechallenge, the preponderance of case reports on second-generation agents is likely explained by confounding factors such as alcohol, cigarettes, and obesity.
Oncology Related Tools
Other
Latest Research
