Introduction:
Pharmacogenetic (PGx) testing has the potential to improve the efficacy and safety of antidepressant pharmacotherapy for moderate-severe major depressive disorder (MDD) by identifying genetic variations that influence medication metabolism and adjusting treatment regimens accordingly.

Aim:
This study aims to assess the cost-effectiveness of implementing a PGx testing approach to guide the prescription of antidepressants.

Methods:
From a public hospital perspective, we developed a two-stage decision tree diagram of a short-term 6-week follow up, and a lifetime Markov model with 3-month cycles. The analysis compared the current standard of care (SoC) with the alternative strategy of PGx-guided (multi-gene panel) testing in adult patients with MDD. Clinical outcomes and utilities were obtained from published studies, while resource costs were from Hamad Medical Corporation, Qatar. The short-term outcome measure was the incremental cost-effectiveness ratio (ICER) against treatment response without side effects (SEs) and without relapse, as well as against treatment response with or without SEs and without relapse. The long-term outcome assessed the ICER against the quality-adjusted life year (QALY) gained and years of life saved.

Results:
Adopting the PGx-guided testing in Qatar resulted in cost savings of Qatari Riyal (QAR) 2,289 (95%CI -22,654-26,340) for the health system. In the short term, the PGx-guided testing was associated with higher response rates without SEs and without relapse (mean difference 0.10, 95%CI 0.09-0.15) and higher response rates with or without SEs and without relapse (mean difference 0.05, 95%CI 0.04-0.06) compared to the SoC. For long term, the PGx-guided testing resulted in 0.13 years of life saved and 0.06 QALYs gained, per person, along with cost savings of QAR 46,215 (95%CI -15,744-101,758). Sensitivity analyses confirmed the robustness of the model results.

Conclusion:
Implementing PGx testing to guide antidepressant use was found to improve population health outcomes, while also significantly reducing health system costs in Qatar.

Introduction: Sepsis, a complex and life-threatening condition, poses considerable challenges in clinical settings due to its high mortality rate. Type 2 Diabetes Mellitus (T2DM) exacerbates sepsis by impairing immune function and altering inflammatory responses. Furthermore, the impact of diabetes on glucocorticoid receptor (GR) functionality exacerbates the severity of the condition.
Aims: We aimed to investigate the association between diabetes and glucocorticoid receptor expression in a cohort of sepsis patients, shedding light on potential implications for the pathogenesis and treatment of sepsis in patients with T2DM.
Methods: Our study involved 47 adult sepsis patients, including 16 with diabetes and 31 without diabetes. Blood samples were collected, and GR expression was quantified using real-time polymerase chain reaction (RT-PCR) from peripheral blood mononuclear cell (PBMC) specimens.
Results: The RT-PCR analysis revealed a difference in GR expression between both groups. Notably, the expression of GRs in the DM group was higher than in the non-T2DM (11.27 vs. 10.85 log 10 gene copies/mL, p= 0.01). Our results also demonstrated that the T2DM group had a higher neutrophil count than the non-DM (p= 0.04). In addition, there were no differences in age distribution between the groups. Furthermore, regarding mortality, the expression of GRs was similar in the patients who experienced mortality compared to the group that survived in the sepsis cohort.
Discussion:
Our investigation revealed that patients with Type 2 Diabetes Mellitus (T2DM) and sepsis exhibited elevated glucocorticoid receptor (GR) protein expression compared to sepsis patients without T2DM. This indicated that diabetic patients may regulate the inflammation caused by T2DM through increased protein production. Targeting GR pathways could offer a promising approach to modulate the immune response and improve outcomes in this population.

-Introduction
In pharmacoepidemiology, to detect early and prevent aggravation, multivariate logistic regression model has been widely applied for disease onset prediction. Tree-based ensemble models which can detect non-linear relationships and handle missing values, are also increasingly applied. Lately, the tree-based ensemble models are highlighted because of their explainability applying Shapley's additive explanatory values (SHAP).

-Aims
To compare the performance of three machine learning models, multivariate logistic regression model (LR), random forests (RF), and gradient boosting decision trees (GBDT), for cardiovascular risk prediction in type2 diabetic patients.

-Methods
The predictive models of cardiovascular events defined by ICD-10 codes, were constructed using a retrospective cohort of type2 diabetic patients (n = 12,301, events = 1,258) extracted from the RWD database between 2008 and 2021. The laboratory test values, including eGFR, creatinine, blood urea nitrogen, were available in the RWD database. LR, RF and GBDT were implemented on Python using LogisticRegression, RandomForestClassifier from scikit-learn and LightGBM Tuner from Optuna, respectively. We compared the performance of the models based on the ROC-AUC scores. Furthermore, the tree-based ensemble models were attempted to explain by means of SHAP.

-Results
The ROC-AUC scores for cardiovascular events in 2 years were 0.5617 with LR, 0.6213 with RF and 0.6464 with GBDT. The tree-based ensemble models showed slightly high performance to LR in this study. SHAP analysis showed that many features, including age, index year, and baseline eGFR, associated with both RF and GBDT. Additionally, SHAP were visualized by force plots to understand the impact of features on the models.

-Discussion
Considering the characteristics of real-world data with many features and missing values, tree-based ensemble models could be an alternative to multivariate logistic regression model in real-world data analysis.

Angiotensin Receptor-Neprilysin Inhibitor (ARNi) is a novel pharmacological class that needs a more comprehensive understanding of the treatment's safety profile.
The study aims to assess the long-term safety of ARNi and ACEi/ARBs after market approval from July 2015 to
August 2021 using electronic medical records. The study also evaluates the difference in hospital length of
stay between the groups, providing a comprehensive evaluation of the clinical outcomes associated with
these drugs.

A retrospective comparative longitudinal cohort study was done using Brigham and Women's Hospital records. Patients were divided into those taking ARNis or ACEi/ARB. To reduce bias, a propensity score matching was conducted. Lab reports were analyzed to compare the incidence of
adverse events (AE) between drug cohorts. We’ve assessed the risks of serum creatinine (Scr>1.4 mg/dl),
hyperkalemia > 5.5 mmol/liter, low systolic blood pressure (SBP< 90 mmHg), as well as patient encounters to
detect angioedema, and acute tubular necrosis. Additionally, HF relatedhospital stay was calculated using Mann-Whitney test .
Results showed that ARNi group had lower incidence of some AE compared to the ACEi/ARB Cohort.
Specifically, the incidence of hyperkalemia (7.7% in the ARNi vs 29.8% in the enalapril ), low systolic blood
pressure (38.4% vs. 47.1%), and high serum creatinine levels (42.2% vs 59.4%) were all statistically significant
lower in the ARNi group. However, the incidence of angioedema and acute tubular necrosis did not show
statistical significance between the two groups.
Additionally, patients in the ARNi group stayed less in the hospital (median = 11 days) than patients in the
ACEi/ARBs group (median = 17days). The Z-score=-6.147,p< 0.001.
These safety analyses showed that the ARNi administration may be associated with a lower risk of certain AE
compared to the ACEi/ARB administration. Additionally, patients treated with ARNi spent less days in the
hospital compared to patients treated with ACEi/ARBs.

INTRODUCTION: The COVID-19 pandemic affected healthcare utilisation because of lockdowns, reductions in non-elective surgeries and stay-at-home orders.

AIMS: This study assessed antibiotic use in Singapore’s public healthcare sector pre- and post-COVID-19 pandemic.

METHODS: We conducted interrupted time-series analyses using aggregated public healthcare institutions’ (PHIs) dispensing data (including hospitals, polyclinics) on oral and parenteral antibiotics, from January 2018 to December 2023. Defined daily doses (DDDs) per 1,000 inpatient days for hospitals and DDDs per 100 doctor visits for polyclinics were calculated and grouped by WHO Access, Watch, Reserve (AWaRe) classification. Access-to-Watch ratio was compared across time periods and care settings.

RESULTS: Overall pre-pandemic utilisation increased by 21.1 in thousand DDDs/quarter (95% CI 1.8, 44.1), mainly in hospitals. In polyclinics, pre-pandemic utilisation decreased by -0.3 DDDs/100 doctor visits (95% CI -0.7, 0.1) but was not significant. In 2020, a significant (p<0.001) and sudden decrease of -437.5K DDDs was observed in PHIs due to lockdowns. After adjusting for inpatient days, the decrease remained significant in the hospitals, especially in the Access antibiotics.

With the rollback of COVID-19 measures from 2021, utilisation increased from 817.8 to 895.8 DDDs/1,000 inpatient days, with a significant trend change of +26.1 (95% CI +15.5, +36.7) in the hospitals. The increase was less pronounced in polyclinics [+1.8 (95% CI +1.2, +2.3)]. Reduction in the Watch antibiotics use (e.g. clarithromycin, ciprofloxacin) post-pandemic led to increase in the Access-to-Watch ratio from 1.77 to 1.95 in the hospitals and 4.25 to 6.73 in polyclinics.

DISCUSSION: Consistent with published studies, the pandemic affected local antibiotic utilisation trends. However, the lockdown effect was more pronounced in the hospitals than polyclinics. The use of Watch antibiotics remains controlled post-pandemic, possibly due to ongoing antimicrobial stewardship in hospitals and prudent use in polyclinics. Future work should examine drivers of antibiotic trend in polyclinics.

Introduction: Recent research has presented divergent conclusions regarding the potential correlation between Sodium-glucose cotransporter 2 inhibitors (SGLT2is) and gastroesophageal reflux disease (GERD).
Aims: The aim of this study was to compare the risk of GERD between SGLT2i users and non-SGLT2i users.
Methods: A cross-sectional study was conducted utilizing data from the 2021 Medical Expenditure Panel Survey (MEPS). We included patients aged over 18, had insurance, diagnosed with type 2 diabetes (T2D), and received at least one anti-diabetes medication. Logistic regressions analyses accounting for complex sampling designs were performed to calculate odds ratios (OR) and confidence intervals (CI). The covariates adjusted in this study were sociodemographics and self-reported health status. Subgroup analyses were carried out to assess differences across various variables, including age, gender, race, ethnicity, education, income, insurance and health status.
Results: The weighted subjects were divided into two groups: 2,799,050 SGLT2i users , and 20,789,424 non-SGLT2i users. Overall, SGLT2i users had higher adjusted odds of GERD compared to non-SGLT2i users (OR=1.64, 95%CI:1.13-2.39). Consistent results were found in the subgroup analyses. Use of SGLT2i was associated with significantly higher odds of GERD in male (OR=2.00, 95%CI:1.21-3.30), black (OR=2.63, 95%CI:1.19-5.82), non-Hispanics (OR=1.49, 95%CI:1.01-2.19), patients with a high school degree (OR=2.14, 95%CI:1.28-3.60), patients with low income (OR=2.11, 95%CI:1.19-3.74), patients with public insurance (OR=2.08, 95%CI:1.27-3.41), and patients with poor to fair health status (OR=2.31, 95%CI:1.32-4.05).
Discussion/Conclusion: Our study found an increased risk of GERD in SGLT2i users, which suggests that SGLT2is should be administered with greater caution among high-risk individuals. However, due to the nature of the cross-sectional design of this study, further studies are needed to clarify the causal relationship between SGLT2i exposure and risk of GERD.

Keywords: sodium-glucose cotransporter 2 inhibitors; gastroesophageal reflux disease; cross-sectional study; Medical Expenditure Panel Survey.

Introduction: Metabolic syndrome (MetS) is a condition caused by lifestyle and with dyslipidemia, type 2 diabetes mellitus, and hypertension. MetS is emphasized because of its association with other serious diseases like stroke. Chronic periodontitis (CP) is also caused by lifestyle and if progresses, affects non-oral diseases. Previous studies have reported interrelation with MetS, remained unclear.

Aims: To examine the association between Mets (dyslipidemia, type 2 diabetes mellitus, and hypertension) and CP in Japan.

Methods: This study used Cross Fact claims database, included from January 2010 to December 2019 and patients who met all following criteria: (1) At least one confirmed diagnosis of dyslipidemia, type 2 diabetes mellitus, or hypertension and (2) 30<= - <65 years old and (3) with 180 days of look back period. The occurrence of CP among MetS patients (%) and characteristics like age at first record of each disease, sex, and the coincidence of other MetS were analyzed.

Results: Patients enrolled in this study were 183,017 (dyslipidemia: 97,498, diabetes: 23,676, hypertension: 107,466 and if diagnosed multiple, counted in each). Patients with CP in this study population were 72,013 (73.9%, 95%CI:73.58-74.14) for dyslipidemia, 16,727 (70.6%, 95%CI:70.06-71.23) for diabetes, 76,654 (71.3%, 95% CI:71.06-71.60) for hypertension, which indicated that CP occurrence were higher in dyslipidemia patients than other two MetS. As characteristics, female showed higher CP occurrence compared to male (75.20%, 95%CI:74.94-75.53 for female, 70.6%, 95%CI:70.3-70.87 for male in study population), which previous studies have reported similarly.

Discussion: Among Mets patients, who had CP showed high proportion regardless of which MetS disease they had. Especially, the proportion of dyslipidemia patients with CP was higher than other two diseases . Further research is required to assess how MetS interact with CP, particularly to explore the factors affect high CP occurrence in dyslipidemia patients.

Keywords: metabolic syndrome, chronic periodontitis, claims database

Introduction: Rheumatoid arthritis (RA) is a chronic inflammatory disorder primarily affecting joints. According to the RA treatment guidelines, Methotrexate (MTX) is the first-line treatment, however, combination therapy with biologic DMARDs or Janus kinase inhibitors (bDMARDs/JAKi) is recommended when MTX alone is insufficient. This study investigates the actual use of anti-rheumatic drugs in patients with RA.

Aims: To investigate the prescription patterns of MTX and bDMARDs/JAKi, and to examine the differences in patient demographic and clinical characteristics of each treatment pattern.

Methods: Using the claims database of Cross-Fact provided by INTAGE Real World Inc., the study included RA patients newly prescribed anti-rheumatic drugs from January 2019 to December 2023. Exclusion criteria included patients with malignancies, immunosuppressive or systemic steroid use, age under 18, certain autoimmune diseases, and a less than 3-month lookback period. The index date was defined by the first date of the prescribed anti-rheumatic drugs. Drug prescriptions and patient demographics were evaluated.

Results: A total of 7,570 eligible patients were included. 3,468 patients were initially prescribed MTX, and 587 were initially prescribed bDMARDs/JAKi. Additionally, 432 were switched to bDMARDs/JAKi from MTX. There was a higher proportion of younger patients aged 18-44 years in both the bDMARDs/JAKi initial prescription group (34.1%) and the bDMARDs/JAKi switching prescription group (33.1%) compared to the other group. There was also a higher proportion of patients treated in hospitals with 20 or more beds in both the bDMARDs/JAKi initial prescription group (68.0%) and the bDMARDs/JAKi switching prescription group (53.5%) compared to the other groups.

Discussion: In the Japanese RA treatment cohort, demographic and clinical factors influence the choice of therapy. Younger patients and those treated in larger hospitals are more often prescribed bDMARDs/JAKi, suggesting early aggressive treatment in this group.

Keywords: rheumatoid arthritis, anti-rheumatic drugs, claims database.

Introduction: Anticholinergic drugs were commonly prescribed for patients with dementia. Although the long-term use of anticholinergic drugs is found to be associated with cognitive impairment in the elderly population, previous studies on the issue often excluded Asians. It is essential to evaluate the prevalence and pattern of anticholinergic drug use among patients with dementia.
Aims: The study aimed to investigate anticholinergic drug utilization patterns among patients with dementia in Taiwan.
Methods: This retrospective study utilized Taiwan’s Longitudinal Health Insurance Database 2005 (LHID 2005), an administrative claims database including health and medication information among residents in Taiwan. The length of the data was from 2005 to 2018. The study population was patients aged over 65, diagnosed with dementia, and received at least one drug for dementia (cholinesterase inhibitors, memantine, and piracetam) for over 30 days. The anticholinergic drugs were categorized into groups of mild, moderate, and severe cognitive impairment based on the Anticholinergic Cognitive Burden scale. Descriptive statistics were used to present the prevalence and pattern of medication use.
Results: There were 11,379 elder patients with dementia enrolled in the study and most of them received piracetam (94.7%) and donepezil (18.2%). In terms of concurrent anticholinergic drug use, drugs with mild cognitive impairment were the most used (49.2%), followed by severe (28.6%) and moderate (22.2%). Specifically, cimetidine (13.8%), fexofenadine (8.3%), and cetirizine (7.7%) ranked as the three most frequently prescribed drugs. Regarding the classification of anticholinergic drugs, those for gastrointestinal conditions and disorders (26.5%) and antihistamines (23.4%) accounted for the majority.
Conclusion: The study showed a high proportion of anticholinergic drug use among older people with dementia in Taiwan. Anticholinergic drugs with mild cognitive impact were reported as the most common use. Future studies could further evaluate the cumulative effect of anticholinergic drug use on dementia among older patients.

Introduction: The Treatment Satisfaction Questionnaire for Medication (TSQM) is a generic measure used to assess treatment satisfaction with medications for patients with a wide range of diseases and modes of administration. All 3 versions of the TSQM have previously been translated into Japanese. However, real-world usage of the measure has identified that the questionnaire includes wording that may be difficult to understand and is also limited to oral medications.
Aims: This study aimed to update the translation of the Japanese version of the TSQM and to reassess the psychometric properties using a sample population of patients with diabetes and migraine in Japan who were prescribed oral and injectable medications.
Methods: The translation and cultural adaptation of the Japanese version of the TSQM were conducted in alignment with the ISPOR task force guidelines. The questionnaire was finalized upon cognitive debriefing interviews with 8 patients. The psychometric properties of the Japanese version of the TSQM were assessed by analysis of Classical Test Theory and Rasch Measurement Theory using data from 512 patients.
Results: Minor modifications were made to the questionnaire following the cognitive debriefing interviews. The results from the psychometric validation found that all Japanese TSQM versions have moderate reliability and high intra-scale validity. Results from the Rasch Measurement Theory added additional support to the reliability and validity of the measures. The scale invariance was supported from differential item functioning across subgroups of patients compared by factors including disease and modes of administration.
Discussion/Conclusion: All Japanese versions of the TSQM were found to have sufficient reliability and validity. These measures were proven to be sufficient regardless of disease or mode of administration. Additional studies investigating the extent of the psychometric properties for modes of administration other than oral or injectable are required.

Introduction
Recently, regulatory authorities worldwide have advised patients receiving statins to be aware of the potential onset of myasthenia gravis (MG) symptoms. Yet, evidence regarding this potential adverse effect were limited to case reports and disproportionality analyses.

Aims
This study aims to examine the risk of incident MG following initiation of statin therapy using multinational real-world data.

Methods
A self-controlled case series (SCCS) study was conducted using electronic medical records and claims databases from Hong Kong, Japan, and United Kingdom. Individuals aged 18 years and above with their first diagnosis of MG and first prescription of any statin during the study period were included. Conditional Poisson regression was employed for within-individual comparison of MG risks during different risk periods (up to two years after statin initiation) compared to the non-exposure period, adjusted for age. Pooled results based on meta-analysis of all study sites were reported.

Results
A total of 2267 MG cases were analysed. Combining all study sites, we observed a significantly increased risk of incident MG during the first year after statin initiation compared to the non-exposure period, with a higher risk during days 0-179 (pooled IRR [95% CI]: 2.662 [1.276-5.553]) than days 180-364 (1.407 [1.014-1.954]). No increased risk of MG was observed more than one year after statin initiation (1.011 [0.848-1.206]). Moreover, the magnitude of MG risk elevation during first 180 days after statin initiation was more pronounced with higher intensity of statins used.

Discussion
In this multinational SCCS study, an increased risk of incident MG during the first 6-12 months after initiation of statin therapy was observed, with greater magnitude of risk elevation for higher intensity statin regimens. Monitoring for incident MG shall be warranted within the first 6-12 months after initiating statin treatment, especially for medium-to-high intensity statin therapy.

Introduction and Aims: We utilized the Korean national health insurance claims database to investigate the potential association between the use of SGLT-2i or DPP4i and the risk of acute pancreatitis through a nationwide case-based analyses.
Methods: Study subjects were diagnosed as type 2 diabetes mellitus who had been prescribed DPP4i, or SGLT-2i at least once between January, 2018 and December, 2020. We defined incident acute pancreatitis cases based on ICD-10 code K85 in either the principal or first additional diagnosis during a hospital admission or emergency department visit. We compared the prescription status of the anti-diabetic medications during the hazard period (30 days before the event of interest), and the preceding control periods of same length with 60 days of washout periods, using conditional logistic regression adjusted for other anti-diabetic medications. When anti-diabetic medication shows a noticeable increase in the exposure trend, we employed the case-case-time-control design to adjust for the effect of the exposure trend on the risk estimate. To ensure the robustness of our findings, a nested case-control study was conducted, with the controls defined within the current type 2 diabetes mellitus cohort.
Results: A total of 7219 incident acute pancreatitis cases were included. In the case-case-time-control analysis, adjusting for the utilization trend, the risk between the acute pancreatitis and SGLT-2i (adjusted odds ratio, 95% confidence interval 0.84, 0.52-1.35) and DPP4i (0.95, 0.54-1.66) was negligible and not statistically significant. The nested case-control study further supported these results, use of SGLT-2i (1.05, 0.91-1.21), DPP4i (1.01, 0.94-1.09) within a 30-day of time window did not show increase the risk of acute pancreatitis.
Conclusion: The use of newly approved hypoglycemic drugs such as SGLT-2i or DPP4i, is not associated with an elevated the risk of acute pancreatitis in patients with type 2 diabetes after considering the trends of anti-diabetic medication prescription.

[Introduction]
While several adverse events (AEs) are known to be associated with polypharmacy, there is a lack of study on the differences by gender.

[Aims]
To investigates gender differences in the safety of polypharmacy among the elderly.

[Methods]
We conducted a retrospective cohort study using the Korean claims data, which includes medical information from one million individuals sampled from 2013-2019. The study population included male and female aged over 65 as of 2014. Patients diagnosed with AE within 2 years prior to index date (January 1, 2015) or those who were polypharmacy patients during the year 2013 were excluded. Polypharmacy was defined as averaging 5 or more prescriptions daily over a year, while non-polypharmacy was defined as averaging fewer than 5 prescriptions daily. The average number of prescriptions per day was calculated by summing the total prescription days for each drug in 2014 and dividing by 365. The five prespecified AEs included hepatic disease, renal disease, fracture, fall, and death, identified by ICD-10 codes. Patients were followed from the date of index date until the earliest diagnosis of AE, death, or end of study (December 31, 2019). We performed 1:4 propensity score matching and estimated incidence rate ratio (IRR) using negative binomial regression model, adjusted for age, residence, Charlson comorbidity index.

[Results]
Out of 93,594 elderly individuals, 42.8% were male, with 14.2% in the polypharmacy group among males. For females, 13.9% were in the polypharmacy group. Hepatic disease (adjusted IRR: 1.11, [95% CI: 1.02-1.20]) showed a statistically significant increase in risk with polypharmacy only in males, while fall (1.80, [1.35-2.41]) and death (2.01, [1.31-3.08]) showed increased risk with polypharmacy only in females.

[Conclusions]
Our research shows that there are differences in polypharmacy safety by gender in hepatic disease, falls, and death. Further gender-specific research is necessary in polypharmacy safety studies.

-Introduction
Pancreatic cancer is a highly aggressive malignancy with a poor prognosis. Current treatment guidelines for metastatic pancreatic cancer emphasize FOLFIRINOX (5-FU, leucovorin, irinotecan, oxaliplatin) or gemcitabine plus nab-paclitaxel as key first-line therapies. However, elderly populations were not the primary focus in clinical trials for these regimens, and comparative studies on their effectiveness and safety in elderly patients are limited.

-Aims
This study aims to evaluate the effectiveness and safety of FOLFIRINOX versus gemcitabine plus nab-paclitaxel in treating elderly patients with metastatic pancreatic cancer.

-Methods
This retrospective cohort study utilized data from the TriNetX Research Network, which includes 88 healthcare organizations. We identified patients aged 65 and older diagnosed with metastatic pancreatic cancer between January 1, 2012, and May 20, 2023, who received either FOLFIRINOX or gemcitabine plus nab-paclitaxel as first-line therapy. The index date was defined as the start date of first-line treatment. Propensity score matching ensured comparability between groups. Survival outcomes were analyzed using the Kaplan-Meier method and Cox proportional hazards models.

-Results
A total of 1,099 patients were identified for the FOLFIRINOX group and 388 for the gemcitabine plus nab-paclitaxel group. After propensity score matching, two cohorts of 360 matched patients each were obtained. The median overall survival was 8.7 months for the FOLFIRINOX group compared to 6.8 months for the gemcitabine plus nab-paclitaxel group, with a hazard ratio of 0.714 (95% confidence interval: 0.598 to 0.851). Incidences of adverse effects included febrile neutropenia (31.4% vs. 19.7%), diarrhea (33.6% vs. 21.7%), and anemia (20.3% vs. 29.2%) for the FOLFIRINOX versus gemcitabine plus nab-paclitaxel groups, respectively.

-Discussion/Conclusion
FOLFIRINOX demonstrated a survival advantage over gemcitabine plus nab-paclitaxel among elderly patients with metastatic pancreatic cancer. However, the higher toxicity associated with FOLFIRINOX necessitates careful consideration by clinicians to balance effectiveness against potential adverse effects.