Abstract
Purpose
Quadruple antiemetic prophylaxis with olanzapine (OLZ) is internationally recommended for cisplatin-based chemotherapy. However, OLZ is contraindicated by country-specific regulatory labeling in patients with diabetes mellitus (DM) in Japan due to concerns regarding hyperglycemia. This study aimed to evaluate the safety of short-term low-dose OLZ (≤10 mg) in patients with cancer and DM.
Patients and Methods
We conducted a retrospective cohort study using the TriNetX multi-institutional electronic medical records network. Patients aged ≥18 years who received cisplatin were included in the study. The primary outcome was the incidence of DM-related adverse events (hyperosmolar hyperglycemic state or diabetic ketoacidosis) within 30 days. Secondary outcomes included glucose elevation, neuropsychiatric events, and acute inpatient visits. Propensity score matching was used to balance the baseline covariates.
Results
In the matched cohorts (n = 84 per group), no DM-related adverse events occurred within 30 days in either group; however, the occurrence of zero events limited definitive comparative inferences. Although patients with DM had higher glucose levels than those without DM at multiple time points, these increases remained modest (mean, approximately 150 mg/dL). No acute inpatient visits or neuropsychiatric events were reported. Sensitivity analyses stratified by the timing and frequency of OLZ exposure confirmed these findings across various exposure patterns.
Conclusion
Based on the absence of detected events rather than demonstrated equivalence, short-term low-dose OLZ administration was not associated with detectable severe glycemic complications in this real-world cohort, supporting prospective evaluation and individualized monitored use in selected patients.
Keywords: chemotherapy-induced nausea and vomiting, antiemetic therapy, hyperglycemia, adverse events, olanzapine
Plain Language Summary
Why was this study done?
Chemotherapy often causes severe nausea and vomiting, which can be effectively prevented by a drug called olanzapine (OLZ). However, OLZ is strictly banned for patients with diabetes in Japan. This is because of concerns that OLZ might dangerously raise blood sugar levels, a side effect typically observed when the drug is used long-term for psychiatric conditions. Researchers wanted to determine whether short-term, low-dose OLZ is actually safe for cancer patients who also have diabetes.
What did the researchers do and find?
The research team examined a large medical database to identify adult patients receiving a chemotherapy drug called cisplatin. They tracked blood sugar levels and severe diabetes-related complications over 30 days and compared diabetic patients who received OLZ with those who did not. The results showed no severe diabetes-related emergencies, such as diabetic ketoacidosis or coma, in any of the patient groups. Although blood sugar levels were slightly higher in the diabetic patients taking OLZ, these increases were modest and manageable.
What do these results mean?
These findings indicate that using low-dose OLZ for a few days does not pose a severe risk to patients with diabetes undergoing chemotherapy. The current strict ban might unnecessarily deprive these patients of highly effective treatment for nausea. The researchers suggest that doctors should be allowed to offer OLZ to diabetic patients alongside regular blood sugar monitoring.
Introduction
Chemotherapy-induced nausea and vomiting (CINV) impairs patients’ quality of life, hindering treatment compliance.1,2 For highly emetogenic chemotherapy, particularly cisplatin (CDDP)-based regimens, international guidelines recommend quadruple antiemetic prophylaxis comprising a 5-hydroxytryptamine (HT3) receptor antagonist, a neurokinin-1 receptor antagonist, dexamethasone (DEX), and olanzapine (OLZ).3–6 OLZ exerts potent antiemetic effects by blocking multiple neurotransmitter receptors, including dopamine (D2), serotonin (5-HT2c, 5-HT3), and histamine (H1) receptors,5,7 exhibiting a crucial role in controlling delayed nausea.8 OLZ is also associated with adverse metabolic events. OLZ-induced hyperglycemia is thought to occur through mechanisms such as receptor antagonism (M3, 5-HT2c, H1) and endoplasmic reticulum stress, which impair insulin secretion and exacerbate insulin resistance.9,10 Long-term administration of OLZ for psychiatric disorders is associated with weight gain and the onset of diabetes mellitus (DM).5 In Japan, OLZ use, according to its package insert, is strictly contraindicated in all patients with DM or those with a history of DM, regardless of their glycemic control. This serves as a regulatory precaution rather than an absolute evidence-based harm for short-term use.8 Conversely, the United States Food and Drug Administration, European Medicines Agency, and international clinical guidelines from the American Society of Clinical Oncology and National Comprehensive Cancer Network [NCCN]), do not list DM as a contraindication. Although these international bodies allow OLZ use under cautious monitoring, Japan maintains strict contraindications based on previous reports of severe hyperglycemia, diabetic ketoacidosis, and hyperosmolar hyperglycemic states.11 This creates a discrepancy between the international standards of care and local regulatory requirements in Japan. Although OLZ is typically administered long-term in psychiatric practice, its use as an antiemetic is generally limited to low doses (5–10 mg) for a short duration of 4 days or on demand. Most pivotal clinical trials for CINV have excluded patients with DM;8,12 therefore, evidence regarding glycemic variability and safety in patients with DM and cancer receiving short-term OLZ during CDDP-based chemotherapy is lacking. When OLZ is omitted from standard antiemetic regimens, a significant proportion of patients experience CINV treatment failure, with complete response rates in the delayed phase being approximately 13–20.5% lower than those in quadruple regimens.8,12 If the risks associated with short-term use are manageable, patients with DM and cancer who are currently denied this antiemetic option could benefit from significantly improved CINV control. This study aimed to evaluate the safety of OLZ in patients with DM and cancer undergoing CDDP-based chemotherapy. We compared glycemic changes, incidence of DM-related complications, and other adverse events (AEs) between patients who received OLZ as an antiemetic and those who did not, to clarify the risk profile of this specific population.
Materials and Methods
Study Design and Data Source
We conducted a retrospective cohort study using the TriNetX Research Network, a multicenter nationwide database providing electronic medical records (EMRs) of approximately 250 million patients.13,14 The data accessed through the TriNetX network are fully de-identified and comply with relevant data protection and privacy regulations, including the Health Insurance Portability and Accountability Act (HIPAA) Privacy Rule. Diagnoses were identified using the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) codes and medications using RxNorm (Table S1). The Natural Language Processing function within TriNetX was used to extract clinical information from unstructured notes. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. We confirmed that no overlapping cohorts from this database have been previously reported by our research group.
Eligibility Criteria
Adults (aged ≥18 years) who underwent their first documented cycle of CDDP were eligible to ensure independent observations. OLZ use was categorized as follows: (1) administration before day 1 to 7 days after CDDP (day −1 to day +7). This broad window was selected to capture both prophylactic use during chemotherapy and rescue use for delayed CINV, allowing a comprehensive evaluation of short-term exposure safety; (2) administration before day 1 to 1 day after CDDP (day −1 to day +1), with no restriction on dosing frequency; and (3) at least three administrations of OLZ between day −1 and day +7.
Category (1) was the primary exposure definition. Categories (2) and (3) were evaluated using a pre-specified sensitivity analysis. DM was defined using ICD-10-CM codes E08–E13 based on diagnoses recorded within 3 years before the CDDP index date.15 We excluded patients with a history of OLZ use, diagnosis of schizophrenia (ICD-10-CM codes F20–F29), or bipolar disorder (F31) within 3 years before the CDDP index date.16
Outcomes and Follow-Up
The primary outcome was the occurrence of DM-related AEs, hyperosmolarity/coma, or ketoacidosis, identified by ICD-10 codes E08.0, E08.1, E09.0, E09.1, E10.0, E10.1, E11.0, E11.1, E12.0, E12.1, E13.0, and E13.1, within 30 days following CDDP administration. Secondary outcomes included neuropsychiatric and neuromuscular events, such as rhabdomyolysis, movement disorders, neuroleptic malignant syndrome, akathisia, convulsions, and unscheduled hospital admissions (Table S2).17 Baseline glucose levels were defined as the most recent measurement obtained within the 30-day period before CDDP administration (day −30 to day −1), and daily glucose values were obtained through day 14. Patients with outcomes before the start of the assessment window were excluded. OLZ administration data were obtained from provider administration records, pharmacy dispensing data, medication reconciliation documentation, and patient history.
Covariates
Covariates (Table S1) included demographic variables (age, sex, and race or ethnicity), hypertension, chronic kidney disease, body mass index (BMI), and glucocorticoid exposure, and malignancies of the respiratory or intrathoracic and digestive organs, with the latter representing common indications for CDDP and potential confounders.18–20 Subgroup analyses included DM-related factors (insulin use, baseline glucose, and oral hypoglycemic agents).
Statistical Analysis
Continuous variables were summarized as means and standard deviations (SDs); categorical variables as frequencies and percentages. Propensity scores were estimated using logistic regression models. Analysis (1) included all covariates. Analyses (2) and (3) included selected covariates (age, sex, malignancy type, and glucocorticoid use) to reduce the risk of overfitting due to the limited sample size. After propensity score estimation, 1:1 greedy nearest-neighbor matching without replacement was performed using a caliper width of 0.2 SDs of the logit of the propensity score. Covariate balance pre- and post-matching was assessed using absolute standardized mean differences (SMDs), with an SMD < 0.1 considered indicative of acceptable balance. Student’s t-tests were used to compare normally distributed continuous outcomes, and chi-square or Fisher’s exact tests were used for categorical outcomes. All statistical analyses were two-sided, and a p value <0.05 was considered statistically significant. Sensitivity analyses were conducted to assess the robustness of the findings under different OLZ exposure definitions. Glucose values were analyzed using available-case analysis at each time point without imputation for missing data. No imputation was performed because the analyses were considered exploratory, and measurement availability varied across time points. Analyses at each time window were conducted cross-sectionally and did not account for within-patient longitudinal correlations. Because glucose distributions are typically right-skewed, blood glucose values were natural log-transformed prior to analysis. All statistical analyses were conducted on a log-transformed scale, and summary statistics were back-transformed to the original scale for interpretability. P-values were adjusted for multiple comparisons using the Benjamini–Hochberg false discovery rate procedure (α = 0.05), and effect sizes (Cohen’s d) were calculated on a log-transformed scale. Statistical comparisons were not performed for outcomes with zero events in either group. The exact 95% confidence intervals for event proportions were calculated using the Clopper–Pearson method. Cohort identification and statistical analyses were conducted on December 4, 2025, using the TriNetX Platform Query Builder and Analytics Functions. Institutional review board approval and informed consent were not required as the study was designated as non–human participant research by the Fukuoka University Institutional Review Board.
Results
Comparison of Patients with and without DM Receiving CDDP and OLZ (Exposure Window: Day −1 to Day +7)
Patient Characteristics
A flow diagram of patients with and without DM receiving CDDP plus OLZ is shown in Figure 1. Among the patients receiving CDDP with OLZ, 90 had a documented history of DM within the preceding year and 614 did not (Table 1). Before propensity score matching (PSM), patients with DM were older, had higher BMI, and included a lower proportion of females compared with those without DM. Cancer type and glucocorticoid use were generally similar in the two groups. After 1:1 PSM, 84 patients remained in each group, with most covariates well balanced between groups.
Figure 1.
Flowchart of patients with and without DM receiving CDDP plus OLZ.
Abbreviations: DM, diabetes mellitus; CDDP, cisplatin; OLZ, olanzapine; PSM, propensity score matching.
Table 1.
Characteristics of Patients with and Without Diabetes Mellitus Among Those Receiving Cisplatin with Olanzapine (Exposure Window: Day −1 to Day +7)
| Characteristic Name | Before PSM | After PSM | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| DM, Yes | DM, No | SMD | DM, Yes | DM, No | SMD | |||||
| n (%) | Mean±SD | n (%) | Mean±SD | n (%) | Mean±SD | n (%) | Mean±SD | |||
| Demographics | ||||||||||
| Age, years | 90 (100.0) | 61.3±10.9 | 614 (100.0) | 54.6±15.5 | 0.504 | 84 (100.0) | 61.1±11.2 | 84 (100.0) | 59.7±11.4 | 0.130 |
| Female | 29 (32.2) | 286 (46.6) | 0.297 | 28 (33.3) | 30 (35.7) | 0.050 | ||||
| BMI, kg/m2 | 74 (82.2) | 30.5±7.5 | 451 (73.5) | 27.0±6.2 | 0.512 | 69 (82.1) | 30.4±7.7 | 70 (83.3) | 29.9±6.5 | 0.075 |
| Race/ethnicity | ||||||||||
| White | 66 (73.3) | 396 (64.5) | 0.192 | 61 (72.6) | 61 (72.6) | <0.001 | ||||
| Hispanic or Latino | 10 (11.1) | 49 (8.0) | 0.107 | 10 (11.9) | 10 (11.9) | <0.001 | ||||
| Black or African American | 12 (13.3) | 57 (9.3) | 0.128 | 11 (13.1) | 11 (13.1) | <0.001 | ||||
| Comorbidities | ||||||||||
| Hypertensive diseases | 67 (74.4) | 197 (32.1) | 0.938 | 61 (72.6) | 60 (71.4) | 0.027 | ||||
| Chronic kidney disease | 12 (13.3) | 18 (2.9) | 0.388 | 10 (11.9) | 10 (11.9) | <0.001 | ||||
| Malignant neoplasms | ||||||||||
| Respiratory and intrathoracic organs | 25 (27.8) | 161 (26.2) | 0.035 | 25 (29.8) | 27 (32.1) | 0.052 | ||||
| Digestive organs | 17 (18.9) | 81 (13.2) | 0.156 | 13 (15.5) | 11 (13.1) | 0.068 | ||||
| Medication | ||||||||||
| Glucocorticoids | 88 (97.8) | 569 (92.7) | 0.241 | 82 (97.6) | 82 (97.6) | <0.001 | ||||
Note: Exposure to olanzapine was defined as ≥1 administration from Day −1 to Day +7 relative to cisplatin (index date = Day 0).
Abbreviations: BMI, body mass index; DM, diabetes mellitus; PSM, propensity score matching; SMD, standardized mean difference; SD, standard deviation.
Clinical Outcomes
Among the propensity score–matched patients receiving CDDP and OLZ (n = 84 per group), blood glucose levels were consistently higher in patients with DM than in those without DM. At baseline (day −30 to −1), the mean glucose level was significantly higher in the DM group (166.4 ± 77.6 mg/dL, n = 56) than that in the non-DM group (123.8 ± 51.6 mg/dL, n = 43; p = 0.002). This trend persisted throughout the follow-up period, with significant differences observed at multiple time points (Figure 2).
Figure 2.
Glucose levels among patients receiving CDDP with OLZ (exposure window: Day −1 to Day +7). Time course of blood glucose levels over the 14 days following CDDP administration. The solid and dotted lines represent the DM-positive and DM-negative groups, respectively. Asterisks indicate an FDR-adjusted p<0.01.
Abbreviations: CDDP, cisplatin; DM, diabetes mellitus; FDR, false discovery rate; OLZ, olanzapine.
No DM-related AEs or acute inpatient visits were observed in either group within 30 days of CDDP administration. Furthermore, no patient in either group developed rhabdomyolysis, neuroleptic malignant syndrome, akathisia, convulsions, or movement disorders within 14 days of CDDP administration (Table 2).
Table 2.
Outcomes Among Patients Receiving Cisplatin with Olanzapine (Exposure Window: Day −1 to Day +7); Outcomes Assessed Through Day 30
| Day | DM, Yes | DM, No | RR (95% CI) | p-value | |
|---|---|---|---|---|---|
| Diabetes-related AE | 30 | 0/84, 0.0%a | 0/84, 0.0%a | – | – |
| Acute inpatient visit | 30 | 0/84, 0.0%a | 0/84, 0.0%a | – | – |
| Neuropsychiatric and neuromuscular AEs | |||||
| Rhabdomyolysis | 14 | 0/84, 0.0%a | 0/84, 0.0%a | – | – |
| Movement disorder | 14 | 0/84, 0.0%a | 0/84, 0.0%a | – | – |
| Malignant neuroleptic syndrome | 14 | 0/84, 0.0%a | 0/84, 0.0%a | – | – |
| Akathisia | 14 | 0/84, 0.0%a | 0/84, 0.0%a | – | – |
| Convulsions | 14 | 0/84, 0.0%a | 0/84, 0.0%a | – | – |
Notes: Adverse events were defined as diagnoses newly recorded after the index date. The olanzapine exposure window was used to define the cohort and did not restrict the follow-up period for the outcome/laboratory assessment. a Zero-event outcomes: 95% exact confidence interval (Clopper–Pearson method), 0.0–4.3%.
Abbreviations: AE, adverse events; CI, confidence interval; DM, diabetes mellitus; RR, relative risk.
Comparison of Patients with and without OLZ Therapy Among Those Receiving CDDP with DM (Exposure Window: Day −1 to Day +7)
Patient Characteristics
A flow diagram of patients with and without OLZ therapy among those receiving CDDP with DM is shown in Figure 3. Among the patients with DM receiving CDDP, 89 were treated with OLZ and 16,671 were not (Table 3). Hypertension (67 [75.3%] vs 11,649 [69.9%]; SMD = 0.121) and chronic kidney disease (13 [14.6%] vs 1702 [10.2%]; SMD = 0.134) were slightly more common in the OLZ group. The prevalence of malignant neoplasms was comparable between the groups. Glucocorticoid use was higher in both groups (88 [98.9%] vs 15,715 [94.3%]; SMD = 0.255). Insulin use was more common in the OLZ group (60 [67.4%] vs 8470 [50.8%]; SMD = 0.343), whereas oral hypoglycemic agents were used at similar frequencies. After 1:1 PSM, 89 patients remained in each group, and the baseline covariates were generally well balanced. Age, sex, BMI, comorbidities, and cancer types were similar between the groups. Glucocorticoid use remained the same in both groups (88 [98.9%]; SMD < 0.001).
Figure 3.
Flowchart of patients with and without OLZ therapy among those with DM receiving CDDP.
Abbreviations: CDDP, cisplatin; DM, diabetes mellitus; OLZ, olanzapine; PSM, propensity score matching.
Table 3.
Characteristics of Patients with and Without Olanzapine Therapy Among Cisplatin-Treated Patients with Diabetes Mellitus (Exposure Window: Day −1 to Day +7)
| Characteristic Name | Before PSM | After PSM | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Olanzapine, Yes | Olanzapine, No | SMD | Olanzapine, Yes | Olanzapine, No | SMD | |||||
| n (%) | Mean±SD | n (%) | Mean±SD | n (%) | Mean±SD | n (%) | Mean±SD | |||
| Demographics | ||||||||||
| Age, years | 89 (100.0) | 60.8±10.2 | 16,671 (100.0) | 63.4±10.2 | 0.252 | 89 (100.0) | 60.8±10.2 | 89 (100.0) | 61.1±11.0 | 0.025 |
| Female | 29 (32.6) | 6419 (38.5) | 0.124 | 29 (32.6) | 31 (34.8) | 0.048 | ||||
| BMI | 71 (79.8) | 30.0±7.4 | 10,460 (62.7) | 29.8±7.1 | 0.018 | 71 (79.8) | 30.0±7.4 | 72 (80.9) | 31.6±8.9 | 0.059 |
| Race/ethnicity | ||||||||||
| White | 68 (76.4) | 9398 (56.4) | 0.434 | 68 (76.4) | 71 (77.1) | 0.076 | ||||
| Hispanic or Latino | 10 (11.2) | 1127 (6.8) | 0.157 | 10 (11.2) | 10 (11.2) | <0.001 | ||||
| Black or African American | 10 (11.2) | 1800 (10.8) | 0.014 | 10 (11.2) | 10 (11.2) | <0.001 | ||||
| Comorbidities | ||||||||||
| Hypertensive diseases | 67 (75.3) | 11,649 (69.9) | 0.121 | 67 (75.3) | 66 (74.2) | 0.026 | ||||
| Chronic kidney disease | 13 (14.6) | 1702 (10.2) | 0.134 | 13 (14.6) | 13 (14.6) | <0.001 | ||||
| Malignant neoplasms | ||||||||||
| Respiratory and intrathoracic organs | 26 (29.2) | 4025 (24.1) | 0.115 | 26 (29.2) | 23 (25.8) | 0.076 | ||||
| Digestive organs | 16 (18.0) | 4309 (25.9) | 0.191 | 16 (18.0) | 15 (16.9) | 0.030 | ||||
| Medication | ||||||||||
| Glucocorticoids | 88 (98.9) | 15,715 (94.3) | 0.255 | 88 (98.9) | 88 (98.9) | <0.001 | ||||
| Insulin | 60 (67.4) | 8470 (50.8) | 0.343 | 60 (67.4) | 58 (65.2) | 0.048 | ||||
| Oral hypoglycemic agents, oral | 42 (47.2) | 7754 (46.5) | 0.014 | 42 (47.2) | 43 (48.3) | 0.022 | ||||
| Hypoglycemic agents, other | 10 (11.2) | 633 (3.8) | 0.285 | 10 (11.2) | 10 (11.2) | <0.001 | ||||
| Laboratory values | ||||||||||
| Glucose | 84 (94.4) | 162.0±68.4 | 13,963 (83.8) | 156.7±68.4 | 0.077 | 84 (94.4) | 162.0±68.4 | 87 (97.8) | 159.7±69.1 | 0.033 |
Note: Exposure to olanzapine was defined as ≥1 administration from Day −1 to Day +7 relative to cisplatin (index date = Day 0).
Abbreviations: BMI, body mass index; PSM, propensity score matching; SMD, standardized mean difference; SD, standard deviation.
Clinical Outcomes
Among the patients with DM receiving CDDP (n = 89 per group), no DM-related AEs occurred within 30 days in either the OLZ or non-OLZ group (Table 4). Mean blood glucose levels were higher in the OLZ group across time points, although not significantly, except for day 5 (p = 0.024). These findings suggest that OLZ use is not associated with clinically meaningful or sustained elevation in blood glucose levels (Figure 4).
Table 4.
Outcomes Among Patients with Diabetes Mellitus Receiving Cisplatin (Exposure Window: Day −1 to Day +7); Outcomes Assessed Through Day 30
| Day | Olanzapine, Yes | Olanzapine, No | RR (95% CI) | p-value | |
|---|---|---|---|---|---|
| Diabetes-related AE | 30 | 0/89, 0.0%a | 0/89, 0.0%a | – | – |
| Acute inpatient visit | 30 | 0/89, 0.0%a | 0/89, 0.0%a | – | – |
| Neuropsychiatric and neuromuscular AEs | 0/89, 0.0%a | 0/89, 0.0%a | – | – | |
| Rhabdomyolysis | 14 | 0/89, 0.0%a | 0/89, 0.0%a | – | – |
| Movement disorder | 14 | 0/89, 0.0%a | 0/89, 0.0%a | – | – |
| Malignant neuroleptic syndrome | 14 | 0/89, 0.0%a | 0/89, 0.0%a | – | – |
| Akathisia | 14 | 0/89, 0.0%a | 0/89, 0.0%a | – | – |
| Convulsions | 14 | 0/89, 0.0%a | 0/89, 0.0%a | – | – |
Notes: Adverse events were defined as newly recorded diagnoses after the index date. The olanzapine exposure window was used to define the cohort and does not restrict the follow-up period for outcome/laboratory assessment. a Zero-event outcomes: 95% exact confidence interval (Clopper–Pearson method), 0.0–4.1%.
Abbreviations: AE, adverse events; CI, confidence interval; DM, diabetes mellitus; RR, relative risk.
Figure 4.
Glucose levels among patients with DM receiving CDDP (exposure window: Day −1 to Day +7). Time course of blood glucose levels over the 14 days following CDDP administration. The solid and dotted lines represent the OLZ-treated and OLZ-untreated groups, respectively. Asterisks indicate an FDR-adjusted p<0.01. Abbreviations: CDDP, cisplatin; DM, diabetes mellitus; FDR, false discovery rate; OLZ, olanzapine.
Sensitivity Analyses Based on OLZ Exposure Timing and Frequency
OLZ Use from Day −1 to Day +1 (Any Frequency)
Table S3 summarizes the baseline characteristics of patients who received OLZ between days −1 and +1. Glucose levels were significantly higher in patients with DM than in those without DM across all evaluated time points (Table S4a). At baseline (day −30 to −1), the mean glucose level was 151.2 ± 40.0 mg/dL in the DM group versus 109.6 ± 28.9 mg/dL in the non-DM group (p = 0.006). Although the difference remained significant at day 1 (DM, 215.5 ± 123.5 vs non-DM, 118.5 ± 25.8 mg/dL; p = 0.026), days 2–8 (170.0 ± 76.6 vs 112.2 ± 20.3 mg/dL; p = 0.020), and days 2–14 (198.1 ± 58.8 vs 113.2 ± 18.4 mg/dL; p < 0.001), no DM-related AEs, inpatient visits, or neuropsychiatric/neuromuscular events occurred in all groups (Table S4b).
Table S5 presents the matched cohort of patients with DM who received CDDP with or without OLZ (n = 20 per group). Baseline glucose levels were similar between the groups (160.1±49.1 vs 151.7±68.0 mg/dL; p = 0.731; Table S6a). Glucose levels tended to be higher in the OLZ group at several time points, although the difference was not statistically significant. No DM-related AEs, inpatient visits, or neuropsychiatric or neuromuscular events were observed (Table S6b).
OLZ Use ≤3 Times from Day −1 to Day +7
Table S7 presents the baseline characteristics of patients who received OLZ ≤ 3 times within the extended exposure window. Among patients with and without DM (n = 31 per group), baseline glucose was significantly higher in the DM group (DM, 144.1 ± 66.3 vs non-DM, 108.6 ± 30.4 mg/dL; p = 0.025), and this difference remained statistically significant at day 1 (185.9 ± 104.1 vs 103.9 ± 25.0 mg/dL; p = 0.003), day 3 (177.7 ± 64.9 vs 131.8 ± 25.7 mg/dL; p = 0.045), day 4 (152.6 ± 45.9 vs 118.8 ± 32.4 mg/dL; p = 0.048), day 7 (136.1 ± 45.8 vs 102.4 ± 28.7 mg/dL; p = 0.034), and day 9 (146.5 ± 56.0 vs 104.4 ± 20.2 mg/dL; p = 0.022) (Table S8a). No DM-related AEs, inpatient visits, or neuropsychiatric or neuromuscular events were observed (Table S8b).
Among patients with DM treated with or without OLZ (n = 33 per group), baseline glucose levels were similar (144.1 ± 66.3 vs 140.3 ± 70.8 mg/dL; p = 0.055) (Table S9). Glucose levels tended to be higher in the OLZ group at day 1 (202.2 ± 97.9 vs 143.0 ± 53.3 mg/dL; p = 0.051), day 2 (182.7 ± 74.3 vs 146.7 ± 52.5 mg/dL; p = 0.183), day 3 (174.9 ± 61.5 vs 151.0 ± 67.1 mg/dL; p = 0.355), and day 4 (153.0 ± 43.1 vs 115.0 ± 28.4 mg/dL; p = 0.020); however, most differences were not statistically significant (Table S10a). No DM-related AEs, inpatient visits, or neuropsychiatric or neuromuscular events were observed (Table S10b).
Discussion
In our study, no severe hyperglycemia or DM-related AE were observed among patients treated with CDDP and OLZ, regardless of their DM status. Among patients receiving CDDP and OLZ, blood glucose levels were significantly higher in those with DM at multiple time points; however, these elevations were modest. Furthermore, no acute inpatient visits or OLZ-related AEs were observed in either group within 30 days of CDDP administration. Consequently, no significant clinical risks associated with short-term OLZ administration in patients with DM were identified.
Sensitivity analyses stratified by the timing and frequency of OLZ exposure corroborated these findings across various exposure patterns. Although a transient increase in blood glucose levels was observed on day 1 of CDDP administration in the OLZ group, the mean values throughout the observation period remained <200 mg/dL. These results suggest that OLZ does not substantially exacerbate preexisting DM. Notably, no serious clinical AEs, such as diabetic hyperosmolarity, coma, or ketoacidosis, were observed. This indicates that glycemic fluctuations associated with short-term low-dose OLZ (≤10 mg) are modest and manageable within the scope of standard cancer care.
The study population consisted predominantly of Caucasian patients and did not include Japanese individuals. Although Japanese-specific inactivating variants of CYP1A2, one of the metabolic enzymes responsible for OLZ biotransformation, have been identified, the metabolic redundancy provided by glucuronidation is considered to preclude any clinically significant interethnic pharmacokinetic differences in the overall clearance or plasma concentrations between Japanese and Caucasian populations.21 Indeed, multiple comparative studies conducted in healthy adult East Asian (Japanese, Chinese, and Taiwanese) and Caucasian subjects have consistently demonstrated the absence of clinically significant interethnic differences in the pharmacokinetics of OLZ.22,23 Accordingly, the findings of the present study are also applicable to the Japanese population.
Current concerns regarding OLZ-induced severe hyperglycemia are based on data derived from its use in psychiatric disorders, such as schizophrenia and bipolar disorder.11 Therefore, information regarding the risks associated with short-term low-dose antiemetic use is limited. OLZ is typically administered at doses of 10–20 mg over a period of several months for the treatment of schizophrenia and bipolar disorder. Conversely, as an antiemetic, OLZ is administered at lower doses (5–10 mg) intermittently every 3 or 4 weeks, typically for 4 days per cycle, or as prophylaxis or rescue medication.3–6 Given OLZ’s half-life, steady-state plasma levels require approximately 7 days,11,24,25 and continuous daily administration data indicate that levels on days 4 and 6 remain below steady state.11,25 Therefore, when administered as an antiemetic, plasma concentrations are anticipated to be lower. Pivotal studies7,8,12 evaluating the efficacy of 5–10 mg OLZ for 4 days reported a low incidence of hyperglycemia, with no significant difference attributable to OLZ administration. In this study, blood glucose elevation in the OLZ group was mild, with no cases of new-onset DM. OLZ addition did not further impair glycemic control, although nearly all participants received corticosteroids. These findings indicate that OLZ does not pose a significant additional clinical concern in patients with DM. While DEX is also associated with hyperglycemia, the NCCN guidelines suggest OLZ as an alternative for patients intolerant to DEX.5 In clinical oncology, patients with impaired glucose tolerance are typically evaluated by endocrinologists before chemotherapy,26,27 and treatment is initiated only if hyperglycemia is managed through caloric adjustment, corticosteroid dose modifications, glucose monitoring, and insulin therapy as needed has improved or is deemed controllable.27,28 In clinical practice, DEX, which carries the risk of hyperglycemia, has been used for many years in patients with DM, and management strategies for its AEs, including glycemic control, are well established.29 Considering the available evidence, many antipsychotics possess hyperglycemic properties. Although diabetic ketoacidosis and hyperosmolar hyperglycemic states are known AEs that require vigilance, their incidence is rare and not unique to OLZ administration.30 Furthermore, no clear signs of OLZ-related AEs (including hyperglycemia) or increased short-term mortality were detected in association with OLZ use in patients with cancer and DM. Thus, the risk of AEs associated with OLZ administration in this patient population appeared potentially manageable, although our finding of zero events in a modest retrospective cohort does not equal confirmed safety. Poor control of nausea and vomiting has been shown to adversely affect survival outcomes31 and quality of life.1 In pivotal clinical trials, quadruple antiemetic prophylaxis incorporating OLZ improved the complete response rate in the delayed phase by 13% in CDDP-based regimens8 and 20.5% in anthracycline plus cyclophosphamide regimens12 compared with triple antiemetic therapy without OLZ. These findings demonstrate the substantial antiemetic efficacy of OLZ. Therefore, overly restrictive approaches to OLZ use in patients with DM may impede the delivery of optimal standard antiemetic therapy to these patients with DM with cancer, potentially limiting access to optimal care. Therefore, a flexible and individualized approach tailored to each patient should be explored.
This study had some limitations. First, the limited sample sizes in several subgroup sensitivity analyses reduced the statistical power to detect rare but clinically meaningful AEs. Although no DM-related or adverse neurological events were observed, these findings should be interpreted with caution. Several serious AEs showed no events in either treatment group. Although exact 95% CI (Clopper–Pearson) were reported, these intervals reflect substantial uncertainty rather than evidence of safety equivalence. Thus, the absence of observed events in a relatively small cohort does not exclude clinically meaningful differences in the true incidence rates between the groups. Larger prospective studies with pre-specified equivalence or non-inferiority margins are therefore needed to establish comparative safety for these rare outcomes. Second, in the PSM analysis, the number of covariates was restricted due to sample size limitations. Although covariates were selected a priori based on clinical judgment and evidence, residual confounding factors cannot be entirely ruled out. To minimize the risk of overfitting, these variables were not included in the propensity score model in certain analyses; however, baseline glucose levels were retained to partially account for the underlying diabetes severity and glycemic control. Third, the study relied on EMR extracted from the TriNetX Research Network; therefore, variability in coding practices, clinical documentation, and medication recordings across institutions may have affected data completeness and accuracy. In addition, the number of patients with available glucose values varied substantially across time points, potentially introducing informative missingness bias. Because of database constraints, we were unable to fully adjust for the dose intensity of dexamethasone, which is a major confounder for glycemia. Furthermore, repeated-measures linear mixed-effects models would have been preferable for accounting for within-patient correlations over time; however, such analyses were not feasible given the database structure and missing longitudinal laboratory data. Fourth, this study focused solely on safety outcomes and could not evaluate the impact of OLZ on the improvement of CINV symptoms. However, given that no marked differences were observed in major patient-related risk factors for CINV, such as age and sex, between our cohort and those enrolled in the aforementioned clinical trials, apart from the comorbidity of DM, OLZ is likely to confer comparable antiemetic benefits in this population. Finally, details regarding the specific intent of OLZ administration (prophylactic or rescue antiemetic) were unavailable. However, by excluding psychiatric indications and restricting our analysis to short-term usage during the peri-administration period of CDDP, we can infer that OLZ (≤10 mg) was prescribed primarily as an antiemetic. We believe that this methodology permits valid risk assessment for short-term low-dose OLZ administration.
Conclusion
Short-term low-dose OLZ was not associated with detectable severe glycemic complications in this real-world cohort, supporting prospective evaluation and individualized monitored use in selected patients.
Acknowledgments
We are grateful to the physicians, nurses, and pharmacists for assisting and caring for the patients. We would also like to express our sincere gratitude to the CINV Committee of the Japanese Association of Supportive Care in Cancer (JASCC) for their valuable advice and helpful discussions. We further acknowledge Editage (Cactus Communications) for the English Language editing of the final draft. Following these services, the authors thoroughly reviewed and edited the manuscript and assumed full responsibility for the content of this publication.
Funding Statement
This work was supported in part by funding from Fukuoka University (Grant No. GR2421).
Abbreviations
5-HT2c, 5-hydroxytryptamine/serotonin receptor 2c; 5-HT3, 5-hydroxytryptamine/serotonin receptor 3; AEs, Adverse events; BMI, Body mass index; CDDP, Cisplatin; CI, Confidence interval; CINV, Chemotherapy-induced nausea and vomiting; D2, Dopamine receptor 2; DEX, Dexamethasone; DM, Diabetes mellitus; EMRs, Electronic medical records; FDR, False discovery rate; H1, Histamine receptor 1; HT3, 5-hydroxytryptamine receptor 3; ICD-10-CM, International Classification of Diseases, Tenth Revision, Clinical Modification; M3, Muscarinic acetylcholine receptor 3 (Implied from “receptor antagonism (M3.)”); NCCN, National Comprehensive Cancer Network; OLZ, Olanzapine; PSM, Propensity score matching; SD, Standard deviation; SMD, Standardized mean difference.
Data Sharing Statement
The data supporting the findings of this study are available from TriNetX; however, restrictions apply to the availability of these data, which were used under the license of the current study and are not publicly available.
Ethics Approval and Informed Consent
Institutional review board approval and informed consent were not required as the study was designated as non–human participant research by the Fukuoka University Institutional Review Board.
Author Contributions
All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising, or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.
Disclosure
The authors report no conflicts of interest in this work.
References
- 1.Sommariva S, Pongiglione B, Tarricone R. Impact of chemotherapy-induced nausea and vomiting on health-related quality of life and resource utilization: a systematic review. Crit Rev Oncol Hematol. 2016;99:13–13. doi: 10.1016/j.critrevonc.2015.12.001 [DOI] [PubMed] [Google Scholar]
- 2.Clark-Snow R, Affronti ML, Rittenberg CN. Chemotherapy-induced nausea and vomiting (CINV) and adherence to antiemetic guidelines: results of a survey of oncology nurses. Support Care Cancer. 2018;26(2):557–564. doi: 10.1007/s00520-017-3866-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Iihara H, Abe M, Wada M, et al. 2023 Japan society of clinical oncology clinical practice guidelines update for antiemesis. Int J Clin Oncol. 2024;29(7):873–888. doi: 10.1007/s10147-024-02535-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Herrstedt J, Clark-Snow R, Ruhlmann CH, et al. 2023 MASCC and ESMO guideline update for the prevention of chemotherapy- and radiotherapy-induced nausea and vomiting. ESMO Open. 2024;9(2):102195. doi: 10.1016/j.esmoop.2023.102195 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.NCCN. NCCN clinical practice guidelines in oncology: antiemesis. version 1. 2026. Available from: https://www.nccn.org/professionals/physician_gls/pdf/antiemesis.pdf. Accessed January 29, 2026.
- 6.Hesketh PJ, Kris MG, Basch E, et al. Antiemetics: ASCO guideline update. J Clin Oncol. 2020;38(24):2782–2797. doi: 10.1200/JCO.20.01296 [DOI] [PubMed] [Google Scholar]
- 7.Navari RM, Qin R, Ruddy KJ, et al. Olanzapine for the prevention of chemotherapy-induced nausea and vomiting. N Engl J Med. 2016;375(2):134–142. doi: 10.1056/NEJMoa1515725 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Hashimoto H, Abe M, Tokuyama O, et al. Olanzapine 5 mg plus standard antiemetic therapy for the prevention of chemotherapy-induced nausea and vomiting (J-Force): a multicentre, randomised, double-blind, placebo-controlled, Phase 3 trial. Lancet Oncol. 2020;21(2):242–249. doi: 10.1016/S1470-2045(19)30678-3 [DOI] [PubMed] [Google Scholar]
- 9.Bymaster FP, Calligaro DO, Falcone JF, et al. Radioreceptor binding profile of the atypical antipsychotic olanzapine. Neuropsychopharmacology. 1996;14(2):87–96. doi: 10.1016/0893-133X(94)00129-N [DOI] [PubMed] [Google Scholar]
- 10.Bajpai J, Kapu V, Rath S, et al. Low-dose versus standard-dose olanzapine with triple antiemetic therapy for prevention of highly emetogenic chemotherapy-induced nausea and vomiting in patients with solid tumours: a single-centre, open-label, non-inferiority, randomised, controlled, phase 3 trial. Lancet Oncol. 2024;25(2):246–254. doi: 10.1016/S1470-2045(23)00628-9 [DOI] [PubMed] [Google Scholar]
- 11.PMDA. Review report—olanzapine. 2017. Available from: https://www.pmda.go.jp/files/000224954.pdf. Accessed December 24, 2025.
- 12.Saito M, Iihara H, Shimokawa M, et al. Overall efficacy and safety of olanzapine 5 mg added to triplet antiemetics for an anthracycline-containing regimen in patients with breast cancer: a phase 3, double-blind, randomised, placebo-controlled trial. Lancet Oncol. 2025;26(7):960–970. doi: 10.1016/S1470-2045(25)00233-5 [DOI] [PubMed] [Google Scholar]
- 13.TriNetX data sets. TriNetX. Available from: https://trinetx.Com/products/real-world-datasets/. Accessed December 4, 2025.
- 14.Palchuk MB, London JW, Perez-Rey D, et al. A global federated real-world data and analytics platform for research. JAMIA Open. 2023;6(2):ooad035. doi: 10.1093/jamiaopen/ooad035 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Abbas S, Ihle P, Köster I, Schubert I. Estimation of disease incidence in claims data dependent on the length of follow-up: a methodological approach. Health Serv Res. 2012;47(2):746–755. doi: 10.1111/j.1475-6773.2011.01325.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Tempelaar W, Kozloff N, Mallia E, Voineskos A, Kurdyak P. Mental health service use before first diagnosis of a psychotic disorder. JAMA Psychiatry. 2024;81(9):928–935. doi: 10.1001/jamapsychiatry.2024.1467 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Sutherland A, Naessens K, Plugge E, et al. Olanzapine for the prevention and treatment of cancer-related nausea and vomiting in adults. Cochrane Database Syst Rev. 2018;9(9):CD012555. doi: 10.1002/14651858.CD012555.pub2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Jones DW, Ferdinand KC, Ferdinand KC, et al; Writing Committee Members. 2025 AHA/ACC/AANP/AAPA/ABC/ACCP/ACPM/AGS/AMA/ASPC/NMA/PCNA/SGIM guideline for the prevention, detection, evaluation and management of high blood pressure in adults: a report of the American college of cardiology/American heart association joint committee on clinical practice guidelines. Circulation. 2025;152(11):e114–e218. doi: 10.1161/CIR.0000000000001356 [DOI] [PubMed] [Google Scholar]
- 19.Parvathareddy VP, Wu J, Thomas SS. Insulin resistance and insulin handling in chronic kidney disease. Compr Physiol. 2023;13(4):5069–5076. doi: 10.1002/cphy.c220019 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.ElSayed NA, McCoy RG, Aleppo G; American Diabetes Association Professional Practice Committee. Diabetes care in the hospital: standards of care in diabetes-2025. Diabetes Care. 2025;48(Suppl 1):S321–S334. doi: 10.2337/dc25-S016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Lin S-K. Racial/ethnic differences in the pharmacokinetics of antipsychotics: focusing on East Asians. J Pers Med. 2022;12(9):1362. doi: 10.3390/jpm12091362 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Kumondai M, Gutiérrez Rico EM, Hishinuma E, et al. Functional characterization of 21 rare allelic CYP1A2 variants identified in a population of 4773 Japanese individuals by assessing phenacetin O-deethylation. J Pers Med. 2021;11(8):690. doi: 10.3390/jpm11080690 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Na Takuathung M, Hanprasertpong N, Teekachunhatean S, Koonrungsesomboon N. Impact of CYP1A2 genetic polymorphisms on pharmacokinetics of antipsychotic drugs: a systematic review and meta-analysis. Acta Psychiatr Scand. 2019;139(1):15–25. doi: 10.1111/acps.12947 [DOI] [PubMed] [Google Scholar]
- 24.Bishara D, Olofinjana O, Sparshatt A, Kapur S, Taylor D, Patel MX. Olanzapine: a systematic review and meta-regression of the relationships between dose, plasma concentration, receptor occupancy, and response. J Clin Psychopharmacol. 2013;33(3):329–335. doi: 10.1097/JCP.0b013e31828b28d5 [DOI] [PubMed] [Google Scholar]
- 25.FDA. Zyprexa (olanzapine) label. 2025. Available from: https://www.accessdata.fda.gov/drugsatfda_docs/label/2025/020592s077,021086s050,021253s047s066lbl.pdf#page=37. Accessed December 24, 2025.
- 26.Gallo M, Muscogiuri G, Felicetti F, et al. Adverse glycaemic effects of cancer therapy: indications for a rational approach to cancer patients with diabetes. Metabolism. 2018;78:141–154. doi: 10.1016/j.metabol.2017.09.013 [DOI] [PubMed] [Google Scholar]
- 27.Societies JBD. The association of British clinical diabetologists. 2021. Available from: https://abcd.care/sites/default/files/site_uploads/JBDS_Guidelines_Archive/JBDS_17_Oncology_Guideline_Final_21052021_Archive.pdf. Accessed December 24, 2025.
- 28.Korytkowski MT, Muniyappa R, Antinori-Lent K, et al. Management of hyperglycemia in hospitalized adult patients in non-critical care settings: an endocrine society clinical practice guideline. J Clin Endocrinol Metab. 2022;107(8):2101–2128. doi: 10.1210/clinem/dgac278 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Barker HL, Morrison D, Llano A, Sainsbury CAR, Jones GC. Practical guide to glucocorticoid induced hyperglycaemia and diabetes. Diabetes Ther. 2023;14(5):937–945. doi: 10.1007/s13300-023-01393-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Polcwiartek C, Vang T, Bruhn CH, Hashemi N, Rosenzweig M, Nielsen J. Diabetic ketoacidosis in patients exposed to antipsychotics: a systematic literature review and analysis of Danish adverse drug event reports. Psychopharmacology. 2016;233(21–22):3663–3672. doi: 10.1007/s00213-016-4411-x [DOI] [PubMed] [Google Scholar]
- 31.Woopen H, Richter R, Chekerov R, et al. Prognostic role of chemotherapy-induced nausea and vomiting in recurrent ovarian cancer patients: results of an individual participant data meta-analysis in 1213. Support Care Cancer. 2020;28(1):73–78. doi: 10.1007/s00520-019-04778-1 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data supporting the findings of this study are available from TriNetX; however, restrictions apply to the availability of these data, which were used under the license of the current study and are not publicly available.