Olanzapine Use and Vomiting Incidence in Chemotherapy-receiving Patients With Cervical Cancer: Multicenter Post Hoc Analysis

Abstract

Background/Aim

Cisplatin administered at a dose of 40 mg/m² in concurrent chemoradiotherapy (CCRT) for cervical cancer is classified as highly emetogenic chemotherapy (HEC), in which the concomitant use of olanzapine is recommended. However, in Japan, the clinical use of olanzapine remains limited owing to pharmacokinetic concerns and contraindications. We aimed to clarify the real-world patterns of olanzapine use and its association with vomiting incidence in patients with cervical cancer undergoing CCRT.

Patients and Methods

This post hoc analysis used data from a multicenter collaborative study of 961 patients with cervical cancer who received CCRT between January 2016 and March 2024. Patient characteristics and antiemetic use, including olanzapine and its dosage, were recorded. Vomiting within six days after the first cisplatin administration was evaluated. Time to vomiting onset was analyzed using Kaplan-Meier methods and multivariate Cox proportional hazards models.

Results

Olanzapine was used in 6.6% overall and 10.6% after guideline revision, indicating limited adoption. Users were younger (median age 52 vs. 56 years) and more frequently received aprepitant and palonosetron. Vomiting incidence was 6.3% vs. 7.6% in users and non-users, with no significant difference, and time to vomiting onset did not differ. Multivariate analysis adjusted for age, neurokinin 1 receptor antagonist use, type of 5-hydroxytryptamine type 3 receptor antagonist, smoking history, and alcohol consumption history showed no significant difference. Among stage IV patients, olanzapine <5 mg/day was more common.

Conclusion

Olanzapine use in patients with cervical cancer undergoing CCRT remained limited and was selectively prescribed to younger high-risk patients. No significant difference in vomiting control was observed, suggesting patient selection bias may have influenced characteristics.

Introduction

Cervical cancer is a gynecological malignancy associated with favorable treatment outcomes when diagnosed at an early stage. In patients with pathological high-risk factors, such as pelvic lymph node metastasis, parametrial invasion, or positive surgical margins, postoperative radiotherapy or concurrent chemoradiotherapy (CCRT) with cisplatin is recommended as adjuvant therapy (1). Cisplatin has a strong emetogenic potential, and doses ≥50 mg/m² are classified as highly emetogenic chemotherapy (HEC) (2). The dose used in CCRT for cervical cancer is 40 mg/m²; however, clinical experience suggests that the addition of radiotherapy-related gastrointestinal toxicity and deterioration in general condition may result in more severe nausea and vomiting than would be expected with conventional moderately emetogenic chemotherapy (MEC) (3). Based on these clinical findings, the Japanese Guidelines for the Appropriate Use of Antiemetics (third edition), revised in October 2023, reclassified cisplatin at a dose of 40 mg/m² used in CCRT for cervical cancer from MEC to HEC. Accordingly, the guidelines now recommend adding olanzapine (days 1-4) and a neurokinin 1 (NK1) receptor antagonist in this treatment setting (4).

However, in Japan, the package insert for olanzapine lists diabetes as a contraindication, and cases of hyperglycemia and new-onset diabetes have been reported (5, 6). Consequently, the uniform administration of olanzapine to all patients is challenging. Furthermore, olanzapine has a prolonged half-life in older patients aged ≥65 years than in younger patients (51.8 vs. 33.8 h), along with reduced clearance (7, 8). Additionally, olanzapine clearance is lower in nonsmokers than in smokers (9). Given these pharmacokinetic properties and underlying disease risks, careful and individualized use of olanzapine is warranted. Moreover, women are known to be at a higher risk of chemotherapy-induced nausea and vomiting (CINV) than men (10). Because all patients with cervical cancer are female, a more individualized antiemetic management strategy is required in this population.

Nevertheless, reports on the real-world use and antiemetic efficacy of olanzapine during CCRT for cervical cancer remain limited. Furthermore, evidence from Japanese patients supporting the recent reclassification of emetogenic risk in the guidelines is scarce, highlighting the need for validation in routine clinical practice. Therefore, in this study, we investigated the real-world use of olanzapine and its association with CINV incidence, with the aim of informing the development of safer and more effective antiemetic strategies for patients with cervical cancer. Accordingly, we focused on vomiting as an objectively documented outcome in real-world clinical records.

Patients and Methods

Study population. This study was a post-hoc analysis of multicenter collaborations using previously reported data (11). Patients with cervical cancer who underwent CCRT between January 1, 2016, and March 31, 2024, were included. Data were collected from 14 institutions across Japan, including university hospitals, cancer centers, and general hospitals. The list of participating institutions and their locations is provided in the authors’ affiliation section. All patients received weekly CCRT consisting of cisplatin at a dose of 40 mg/m².

The exclusion criteria were participation in other clinical or interventional trials, missing data for the study variables, and documented prior use of olanzapine pre-CCRT initiation, which was excluded to avoid potential confounding in the assessment of the antiemetic effect of olanzapine initiated at the time of first cisplatin administration.

Study variables. Patient characteristics included age, smoking history, alcohol consumption history, cancer stage, and prior chemotherapy. Regarding antiemetic use, the presence and type of NK1 and 5-hydroxytryptamine-3 (5-HT3) receptor antagonists, as well as olanzapine use and dosage, were recorded. Cancer stage was determined according to the International Federation of Gynecology and Obstetrics classification, which is commonly used in Japanese gynecologic oncology practice. All variables were assessed based on data obtained at the time of the first cisplatin administration.

For vomiting, the observation period was defined as the six days following the first cisplatin dose. Patients were classified as having experienced vomiting if any episode was documented in the electronic medical records, and the date of the first recorded episode was defined as the day of vomiting onset.

Statistical analysis. Categorical variables were compared between groups using Fisher’s exact test, and continuous variables were compared using the Mann-Whitney U test. Time to vomiting onset within six days after the initial cisplatin administration was analyzed using the Kaplan-Meier method, and differences between groups were assessed using the log-rank test. The numbers at risk shown below the Kaplan-Meier curve represent the number of patients who had not experienced the event immediately before each time point. Younger age has been reported as a risk factor for CINV, whereas NK1 receptor antagonists and palonosetron have been shown to be effective in preventing CINV (10, 12, 13). Additionally, previous studies have identified patient-related factors such as smoking and alcohol consumption as potential risk factors for CINV (14, 15). To adjust for potential confounding by these factors, a multivariate analysis was performed using a Cox proportional hazards model with age, NK1 receptor antagonist use, type of 5-HT3 receptor antagonist, smoking history, and alcohol consumption history as covariates. Because all patients received a 5-HT3 receptor antagonist, adjustment for its use was not performed. All tests were two-sided, and p<0.05 was considered statistically significant. Statistical analyses were conducted using EZR version 1.55 (Saitama Medical Center, Jichi Medical University, Saitama, Japan) (16).

Ethical considerations. This study was conducted in accordance with the Ethical Guidelines for Medical and Life Science Research Involving Human Subjects and was approved by the Ethics Committee of Ehime University Hospital (approval number: 2511005). Approval was also obtained from the ethics committees of all participating institutions prior to data collection. Because this was a retrospective, observational study, it was deemed unnecessary to obtain individual informed consent from the patients. Information about the study was provided on each, institutional website along with an opt-out clause.

Results

Trends in olanzapine usage. A total of 1,074 patients with cervical cancer underwent CCRT during the study period. After excluding patients who met the exclusion criteria (17 participating in clinical trials, 91 with missing data, and five already taking olanzapine at CCRT initiation), 961 patients were included in the analysis, as shown in Figure 1.

Figure 1.

Flowchart of patient selection and dataset construction.

Trends in olanzapine use during the observation period are shown in Figure 2. No patients received olanzapine between 2016 and 2017. Overall, olanzapine was administered to 6.6% (63/961) of patients at the time of the first cisplatin dose. During the six months following guideline revision (October 2023-March 2024), the olanzapine use rate was 10.6% (7/66), compared with 8.5% (25/295) during the pre-revision period (January 2022-September 2023), representing a 2.1% increase.

Figure 2.

Trends in olanzapine usage.

Baseline characteristics of patients with and without olanzapine prescription at the first cisplatin administration. Of the 961 patients included in the analysis, 6.6% (63/961) received olanzapine, while 93.4% (898/961) did not. Baseline characteristics are summarized in Table I. A comparison of baseline characteristics revealed a significant difference in age between the two groups, with patients in the olanzapine group being younger than those in the non-olanzapine group (median age: 52 vs. 56 years, p=0.01).

Table I. Baseline characteristics of patients with and without olanzapine prescription at the first cisplatin administration.

^aMedian (IQR). ^bIncidence of vomiting after the first course administration with a 6-day observation period. ^cMann–Whitney U-test. ^dFisher's exact test. 5-HT3: 5-hydroxytryptamine 3; NK1: neurokinin 1; IQR: interquartile range.

Regarding antiemetic use, the proportion of patients receiving an NK1 receptor antagonist was significantly higher in the olanzapine group than in the non-olanzapine group (p<0.001). Differences were observed in the type of NK1 receptor antagonist used, with aprepitant administered more frequently in the olanzapine group (p=0.048). All patients in both groups received 5-HT3 receptor antagonists, and palonosetron use was significantly higher in the olanzapine group than in the granisetron group (p<0.001).

Vomiting incidence was 6.3% (4/63) in the olanzapine group and 7.6% (68/898) in the non-olanzapine group, with no significant difference (p=1.000).

Patient characteristics according to olanzapine dose at the first cisplatin administration Daily olanzapine doses were 1.5 mg in 1.6% (1/63) of patients, 2.5 mg in 23.8% (15/63), 5 mg in 73.0% (46/63), and 10 mg in 1.6% (1/63). Patients were divided into two groups: <5 mg/day (n=16) and ≥5 mg/day (n=47). Comparison of patient characteristics between these dose groups is summarized in Table II. Stage IV disease was significantly more common in patients receiving <5 mg/day of olanzapine (p=0.006).

Table II. Patient characteristics according to olanzapine dose at the first cisplatin administration.

^aMedian (IQR). ^bMann–Whitney U-test. ^cFisher's exact test. 5-HT3: 5-hydroxytryptamine 3; NK1: neurokinin 1; IQR: interquartile range.

Time-to-event analysis for vomiting according to olanzapine use at the time of the first cisplatin dose. Time to vomiting onset in the olanzapine and non-olanzapine groups is shown in Figure 3. The mean time to vomiting was 5.86 and 5.76 days in the olanzapine and non-olanzapine groups, respectively, with no significant difference (p=0.71). A Cox proportional hazards model adjusting for age, NK1 receptor antagonist use, type of 5-HT3 receptor antagonist, smoking history, and alcohol consumption history revealed no significant difference between groups (hazard ratio=2.56, 95% confidence interval=0.85-7.66, p=0.09).

Figure 3.

Time-to-event analysis for vomiting according to olanzapine use at the time of the first cisplatin dose. Numbers below the curve indicate patients at risk. NK1: Neurokinin 1; 5-HT3: 5-hydroxytryptamine-3; CI: confidence interval.

Discussion

In this multicenter study, we analyzed 961 of 1,074 cervical cancer cases undergoing CCRT, examining real-world olanzapine use and its association with vomiting in Japan. Overall, olanzapine was administered to 6.6% of patients during the observation period, rising to only 10.6% in the six months following guideline revision. This highlights that guideline-recommended antiemetic therapy has not yet been fully implemented. Previous reports indicate that it took approximately two years for NK1 receptor antagonist use to exceed 60% after their inclusion in the Japanese antiemetic guidelines (17), suggesting that the introduction of olanzapine in the CCRT setting may still be in a transitional phase. Additionally, pharmacokinetic concerns, including contraindications in patients with diabetes listed on the package insert (5, 6), prolonged half-life and reduced clearance in older patients (7, 8), and lower clearance in non-smokers (9), along with potential adverse events such as somnolence and metabolic disturbances, likely influenced physicians’ prescribing decisions. Nausea and vomiting during radiotherapy or chemoradiotherapy for cervical cancer have been reported to be common and influenced by multiple clinical factors, indicating that optimal supportive care remains an important issue, and there is still room for improvement in treatment strategies (18, 19).

Olanzapine administration was significantly more frequent in younger patients than in older patients (p=0.01, Table I). Because young women are known to be at higher risk for CINV, olanzapine may have been preferentially prescribed to patients with a greater baseline risk. Additionally, decreased olanzapine clearance has been reported in patients aged ≥65 years (8), which may have led physicians to prescribe olanzapine more cautiously in older patients. Notably, patients with stage IV disease were significantly more likely to receive olanzapine at doses <5 mg/day (p=0.006, Table II), suggesting that dose reductions may have been selected in advanced cases, possibly reflecting overall patient condition. Collectively, these findings suggest that both the increased CINV risk in younger patients and the pharmacokinetic considerations in older patients likely influenced clinical decisions regarding olanzapine administration.

Furthermore, the olanzapine group had higher rates of aprepitant and palonosetron use; however, vomiting incidence after the first cisplatin dose did not differ significantly between the olanzapine and non-olanzapine groups (p=1.00, Table I). In the Cox proportional hazards model, olanzapine administration was not significantly associated with time to vomiting onset (hazard ratio=2.56, 95% confidence interval=0.85-7.66, p=0.09, Figure 3). In addition, in the log-rank analysis, there was a trend toward higher vomiting risk in the olanzapine group, although this was not statistically significant. This may reflect the influence of indication bias, as olanzapine was preferentially administered to patients at higher risk of vomiting. These findings differ from the reported efficacy of aprepitant and palonosetron (10-12) and may reflect insufficient statistical power, given the lack of statistical significance despite an HR exceeding 1. While a prospective randomized controlled trial has shown the efficacy of antiemetic therapy, including olanzapine (18), the results of the present study are not directly comparable. Hashimoto et al. conducted a double-blind, randomized controlled trial to compare patients receiving olanzapine in addition to standard therapy with those not receiving olanzapine among patients undergoing HEC (cisplatin ≥50 mg/m²) and reported a significant reduction in the incidence of delayed-phase vomiting (20). In that trial, CINV risk factors such as age and sex were balanced between groups through randomization, enabling a more precise evaluation of the antiemetic effect of olanzapine. In contrast, the present study was retrospective and observational, and baseline CINV risk may not have been balanced between groups. It is possible that olanzapine was selectively administered to patients at a higher risk of developing CINV. Therefore, the true antiemetic effects of olanzapine may have been underestimated in the present study. Furthermore, chemotherapy-induced nausea and vomiting are associated with decreased food intake and impaired quality of life, and individual risk assessment is important for optimizing antiemetic strategies. In addition, supportive care interventions, as suggested in previous multicenter studies, may also influence vomiting outcomes. Taken together, these considerations warrant careful interpretation of the present findings (21-23).

Study limitations. First, because this was a retrospective observational study, the decision to administer olanzapine was left to individual attending physicians, making selection bias and confounding related to patient characteristics and institutional policies unavoidable. Second, our analysis focused primarily on vomiting, and olanzapine-specific adverse events such as somnolence, dizziness, and fluctuations in blood glucose levels were not adequately evaluated. Furthermore, assessments of nausea and quality of life were not performed. Therefore, further prospective studies are warranted to validate the study findings. Finally, the observation period following the guideline revision was only six months, which is insufficient to capture medium- to long-term trends in olanzapine use or changes in prescription patterns associated with its broader adoption. Consequently, these results reflect a “transitional” state of olanzapine use in Japan and should not be interpreted as defining optimal practice.

Nevertheless, this study is significant in clarifying real-world olanzapine use in a large multicenter cohort of patients undergoing CCRT for cervical cancer. Despite the classification of this regimen as HEC in current guidelines, the clinical use of olanzapine remains limited. When prescribed, olanzapine was preferentially administered to younger patients and those at higher risk of vomiting, and dose reductions below the guideline-recommended levels were observed in advanced-stage cases, reflecting flexible, patient-tailored management. Moving forward, broader dissemination of appropriate antiemetic strategies, supported by up-to-date evidence-based information and educational efforts in clinical practice, will be essential to optimize care for high-risk patients.

Conflicts of Interest

The Authors declare that this study was conducted in the absence of any commercial or financial relationships that could be construed as potential conflicts of interest.

Authors’ Contributions

Yusuke Kawamura and Kensuke Yoshida designed the study and made critical revisions. Yusuke Kawamura was involved in the analysis and writing of the manuscript. Yusuke Kawamura analyzed the data set. Kensuke Yoshida and Sugiura reviewed and edited the manuscript. All the Authors approved the final version of the manuscript.

Acknowledgements

The Authors want to thank Mr. Noriaki Hidaka, Ms. Chie Saito, Mr. Takahiro Kobayashi, Mr. Tomokazu Shoji, Ms. Motoko Kaneko, Mr. Masayoshi Koga, and Mr. Tomoya Nozaki for summarizing the questionnaire data and Editage (www.editage.jp) for English language editing.

Funding

No funding was received for conducting this study.

Artificial Intelligence (AI) Disclosure

No artificial intelligence (AI) tools, including large language models or machine learning software, were used in the preparation, analysis, or presentation of this manuscript.