Incidence of Chemotherapy-Induced Nausea and Vomiting and Antiemetic Guideline Compliance in Chinese Real Clinical Setting: A Cross-Sectional Multicenter Survey

Background This survey aims to investigate the incidence of chemotherapy-induced nausea and vomiting (CINV) in Chinese real clinical setting and evaluate the effect of guideline-consistent CINV prophylaxis (GCCP) and guideline-inconsistent CINV prophylaxis (GICP) on incidence of complete response (CR) of CINV. Materials and Methods A cross-sectional nationwide multicenter study assessing the guideline consistency and CINV incidence of patients was conducted at a total of 32 large medical centers from 26 provinces across the west,east (cid:0) northeast and middle part of China between April and May 2021. Result Data for 2964 patients were analyzed. Patients treated with moderately emetogenic chemotherapy (MEC) were more prone to experience CINV during the acute phase compared to those receiving highly emetogenic chemotherapy (HEC); patients receiving low or minimally emetogenic chemotherapy (L/mEC) were least likely to experience CINV during the overall phase among the whole study population. The prevalence of GCCP was 29.2% in the whole study population, and 13.6%, 35.7% and 45.1% for the patients receiving HEC, MEC and L/mEC, respectively. For patients receiving HEC and MEC, GCCP increased incidence of CR during both delayed and overall phases. For those receiving L/mEC and GICP, incidence of CR was not higher than that of patients receiving L/mEC and GCCP. Conclusion This study revealed Chinese CINV status, the prevalence of GCCP in the real clinical setting and the association between GCCP and CR rate for the rst time. The ndings indicate that prescribing antiemetics in compliance with guidelines for all patients receiving chemotherapy is strongly suggested. shows the Incidence of acute CR between GCCP and GICP in patients receiving HEC, MEC and L/mEC, respectively, and in the whole study population. (C) The graph shows the Incidence of delayed CR between GCCP and GICP in patients receiving HEC, MEC and L/mEC, respectively, and in the whole study population. The data are presented as percentages and were tested using the chi-square test.


Introduction
Chemotherapy-induced nausea and vomiting (CINV) is considered as a common, debilitating reaction to chemotherapeutic agents [1]. The incidence of CINV has been well controlled as the development of effective antiemetics [2]. Although effective antiemetics and consensus guidelines for CINV are available nowadays, CINV still deteriorate patients' quality of life and in the worst case even lead to discontinuation of chemotherapy, compromised effectiveness of chemotherapy and more costs on outpatient visits and hospitalization [3].
According to international guideline, chemotherapy regimens are categorized into four classes based on the antiemetic with the highest emetogenicity in the regimen [4], which are highly emetogenic chemotherapy (HEC), moderately emetogenic chemotherapy (MEC), low emetogenic chemotherapy (LEC), and minimally emetogenic chemotherapy, respectively. CINV could occur in a biphasic mode consisting of acute(≤24h from administration of chemotherapy,day1)phase and delayed >24h from administration of chemotherapy, days 2-5 phase. Previous studies reported that CINV was more prone to occur in the delayed phase as compared with the acute phase [5].
Guideline consistency in practice is variable across different countries, Studies conducted in the United states and Europe have reported the unsatisfactory incidence of guideline adherence which were 57.3% and 29 %, respectively, and the ndings in these two studies indicated that compliance with antiemetic guidelines has a positive effect on control of CINV during the overall phase (acute phase plus delayed phase) for patients administered HEC or MEC [6,7]. Thus the primary objective of this survey was to analyze the impact of guideline-consistent CINV prophylaxis (GCCP) and guideline-inconsistent CINV prophylaxis (GICP) on incidence of complete response (CR) in China. Besides, a number of clinical trials reported that use of newly modern antiemetics could increase rate of complete response (CR), such as palonosetron (PALO) and aprepitant [8,9]. Considering few studies about the effectiveness of 5hydroxytryptamine-3 receptor antagonist (5HT3-RA) or aprepitant in combined regimens in the real clinical setting, investigating their effectiveness in combined regimens in Chinese clinical practice was also conducted.

Materials And Methods
This was a cross-sectional nationwide multicenter survey conducted at 32 large medical centers including cancer centers and general hospitals from 26 provinces across the west,east northeast and middle part of China between April and May 2021. To keep balance, the number of patients at each study site was requested to range from 30 to 120. All methods were carried out in accordance with relevant guidelines and regulations. All experimental protocols were approved by Ethics Committee of Sun Yat-sen University Cancer Center. Informed consent was obtained from all subjects and all subjects are above 18, and the ethical approval was provided in supplement 2.

Patients
Patients who were ≥ 18 years old and had already received HEC, MEC, low or minimally emetogenic chemotherapy (L/mEC) were eligible for the study. Patients were excluded if they had been received opioids or other drugs that interfere with the metabolism of antiemetics during chemotherapy treatment, and those who cannot express objective discomfort, cannot take food because of digestive tract obstruction, with brain metastases and concurrent radiation therapy were also excluded.

Questionnaire design and outcomes
The questionnaire included six sections as follows: demographic information, patients' recall about control of CINV after the most recent chemotherapy, antiemetic regimens, adverse effects of antiemetics, treatment adherence and medical costs. Oncologists lled out the questionnaire via asking patients predesigned questions (Questionnaire is presented in supplemental 1).
Analyzing the data from questionnaire led to following outcomes: demographics, use of antiemetic medications for CINV, incidence of CINV, the effect of GCCP and GICP on incidence of complete response (CR) and prognostic factors for the overall CR. CR referred to no vomiting and no rescue therapy. This study analyzed CR during the acute, delayed and overall phases, respectively. The Emetogenicity of chemotherapy regimen was categorized into four classes,and the de nition of GCCP in this study was based on the NCCN Antiemesis Guidelines Version 1.2019 [4] (supplemental Table 1). In our study, LEC and minimally emetogenic chemotherapy were analyzed as a group named L/mEC. For patients receiving LEC, GCCP was de ned as use of single agent as follows: dexamethasone, metoclopramide, prochlorperazine or 5HT3-RA; using more than one antiemetic for patients receiving LEC was regarded as GICP; for patients receiving minimally emetogenic chemotherapy, no routine prophylaxis was de ned as GCCP.

Statistical Analyses
Demographics and clinical characteristics of patients were compared among HEC,MEC L/mEC groups using Chi-square test for categorical variables and Kruskal-Wallis test for age; post-hoc analysis was used to compare difference in distribution for gender and use of antiemetic therapy. Chi-square test was used to compare the incidence of CR between GCCP and GICP. Prognostic factors for CR during the overall period were identi ed via univariate and multivariate logistic regression analysis. The independent factors included in the analysis are as follows: gender ,age, number of prophylactic antiemetics, use of 5HT3-RA, emetogenic risk level of chemotherapy, use of aprepitant, history of anticipatory nausea and vomiting (ANV), education level, degree of knowing illness and guideline compliance [10][11][12].Two-sided P values < 0.05 were accepted as statistically signi cant. STATA software, version 15.0 was used for all statistical analyses.

Ethical Considerations
This cross-sectional multicenter survey was initiated by the Clinical Chemotherapy Committee of Chinese Anti Cancer Association (CACA) and the contents of questionnaire enshrine ethical considerations, without any invasion of privacy. All methods were carried out in accordance with relevant guidelines and regulations. All experimental protocols were approved by Ethics Committee of Sun Yat-sen University Cancer Center. Informed consent was obtained from all subjects and all subjects are above 18, and the ethical approval was provided in supplement 2.

Patients
Of 3000 patients enrolled, as data for chemotherapy regimen were missing for 36 patients, data for 2964 patients treated with HEC, MEC or L/mEC were analyzed. Table 1 summarizes the characteristics of the 2964 patients (1184 treated with HEC, 1037 treated with MEC and 743 treated with L/mEC; 1516 men and ages in HEC, MEC and L/MEC were 56,56 and 57 years old, respectively. The 3 groups (HEC, MEC, L/mEC) of patients shared an unsimilar distribution of gender, prevalence of antiemetic usage, degree of knowing illness and percentages of cancer type (P < 0.01). General hospital and cancer center showed a similar distribution among 3 groups (P = 0.179). Greater proportions of male patients were in HEC and MEC cohort compared with L/MEC cohort (P < 0.001). A lower usage rate of antiemetics in MEC and L/mEC compared to HEC cohort (P < 0.001).  Table 2). 5HT3-RA nearly became the basis of an antiemetic regimen as 2876 out of 2964 patients had 5HT3-RA in their prophylaxis antiemetic regimens. PALO was the most commonly prescribed 5HT3-RA (n = 1068;37.13%), followed by Tropisetron (n = 817;28.41%), and only 2.68% (n = 77) were administered Azasetron which was the least used 5HT3-RA.
CR outcomes of two speci c antiemetic regimens for HEC and MEC were describes respectively in supplemental Table 3. The patients treated with HEC and aprepitant-containing triple antiemetics (5HT3-RA + dexamethasone + aprepitant) had statistically higher CR rates compared with patients treated with dual antiemetics (5HT3-RA + dexamethasone) during the acute phase and overall phase; but for the patients treated with MEC, the CR rate had no signi cant difference between above-mentioned dual antiemetics and triple antiemetics. Among the patients after HEC and triple antiemetics(5HT3-RA + dexamethasone + aprepitant), the CR rates between patients treated with PALO and patients treated with 1st generation 5HT3-RA did not show a statistically signi cant difference. Regarding the patients after MEC treated with dual antiemetics (5HT3-RA + dexamethasone) or triple antiemetics (5HT3-RA + dexamethasone + aprepitant), the CR rate in patients prescribed PALO was not statistically different from that in patients prescribed 1st generation 5HT3-RA.
During overall, acute and delayed phases, patients were more likely to experience nausea compared with vomiting ( Fig. 1) Fig. 2, for the whole study population, higher incidence of CR in patients treated with GCCP was observed during overall, acute and delayed phases (P = 0.023 for acute CR; P < 0.001 for both delayed and overall CR). GCCP could signi cantly increase the incidence of CR in the patients treated with HEC (P = 0.007 for the delayed CR; P < 0.001 for the acute and overall CR). Regarding the patients treated with MEC, the incidence of acute CR in patients receiving GCCP was not statistically higher than that in the patients receiving GICP (P = 0.311), but CR rates were both statistically higher in the patients treated with GCCP than that in the patients treated with GICP for both delayed and overall phases (P = 0.042,P = 0.007, respectively). For the patients treated with L/mEC, no statistically signi cant difference in CR rates between GCCP and GICP was observed during the acute, delayed and overall phases.
Univariate and Multivariate analysis of prognostic factors for overall CR Univariate and multivariate logistic regression analyses were performed to identify several statistically signi cant predictors for overall CR (  Abbreviation: CI, con dence interval; ANV, Anticipatory nausea and vomiting; HEC,highly emetogenic chemotherapy;MEC, moderately emetogenic chemotherapy; L/MEC, low/minimally emetogenic chemotherapy; OR odds ratio; GCCP guideline-consistent CINV (chemotherapy induced nausea and vomiting) prophylaxis GICP guideline-inconsistent CINV chemotherapy induced nausea and vomiting prophylaxis.

Discussion
For the rst time this study revealed Chinese CINV status, the prevalence of GCCP in the real clinical setting and the association between GCCP and CR rate. The ndings in this study indicate that CR rate in the patients treated with GCCP beyond that seen with patients treated with GICP, thus standardizing the use of antiemetics in real clinical setting is of great signi cance. The strengths of this study lie in many aspects. It consisted of a total of 32 large medical centers and 2964 patients from 26 provinces of China. Compared with the previous western studies [6,7], the bigger sample size in this study provided a more robust evaluation of the bene ts of GCCP. More notably, the incidence of CR after patients receiving L/mEC was compared between GCCP and GICP for the rst time [13], as patients treated with L/mEC were not enrolled in previous western studies. . So ndings in this study support prescribing antiemetics in compliance with guidelines in the real clinical setting for patients receiving HEC, MEC as GCCP could signi cantly increase the incidence of CR for patients receiving HEC and MEC. Besides, for the patients receiving L/mEC, no signi cant difference in CR rate between GCCP and GICP was observed, indicating that excessive treatment (more than one antiemetic) for patients receiving LEC and prescribing antiemetics for patients receiving minimally emetogenic chemotherapy could not increase the incidence of CR.
In this study, only 13.6% of patients prescribed HEC and 35.7% of patients prescribed MEC complied the NCCN 2019 guidelines, respectively, which were comparable with the results reported in the conducted by Aapro et al. (11%, 39%, respectively). However, this was not a satisfactory result when compared with the GCCP prevalence in a Japanese study, in which Hirotoshi et al. reported 96% of patients receiving carboplatin-based chemotherapy (HEC and MEC were both included) were in compliance with the antiemesis guidelines based on JSCO 2010 guidelines [14]. Except the fact of using different guidelines, Japan enacted the rst cancer control law in 2006, which con rmed that political policies could in uence compliance with guidelines [14]. Economic burden is another notable issue. The cost-effectiveness of using aprepitant was reported quite good in Germany, the United States and the United Kingdom [15]. But it has not been involved into Chinese health insurance, which might be responsible for the insu cient use of aprepitant in patients receiving HEC (only 255 out of 1184 patients receiving HEC were administered aprepitant in this survey), hence improving health care policy might be a practical measure to increase guideline consistency and reduce the burden of CINV.
The incidence of nausea during overall, acute and delayed phases was more frequent than vomiting, which was consistent with previous studies [16,17]. Ng, T. L et al. reported that patients ranked nausea more of a problem than vomiting [18]. The incidence of nausea during the overall phase was approximately 3.5-fold higher than that in the vomiting. This implies the control of nausea in real clinical setting was much worse compared to the control of vomiting. For the overall study population, data generated in this study found that the incidence of acute CINV was higher than the incidence of delayed CINV, unlike results reported by previous studies indicating that incidence of delayed CINV was higher [16,20].This might because those previous studies analyzed rst cycle of chemotherapy among chemotherapy-naïve patients [21]. while our study, as a cross-sectional study, referred to patients' most recent treatment cycle consisting of diverse chemotherapy regimens, confounding factors were inevitable. Vidall et al found the same pattern of CINV incidence as our study which might because they also referred to patients' most recent cycle of chemotherapy [22]. The consistency might imply the incidence of CINV in the acute phase was indeed higher compared with delayed phase in real clinical setting. Unexpectedly, higher percentage of patients receiving MEC experienced nausea compared to patients receiving HEC, although the prevalence of GCCP in MEC was higher than that in HEC. This could be attributed to the insu cient education and contacts between oncologists and patients receiving MEC, unlike the patients receiving HEC, who were more prone to be educated about the prophylaxis of CINV.
5HT3-RA was the basis of an antiemetic regimen as 97.03% of patients had 5HT3-RA in their prophylaxis antiemetic regimens. A phase III clinical study reported that among different types of 5HT3-RA, PALO as the most commonly prescribed, its complete response (CR) rates for CINV were signi cantly higher compared with 1st generation 5HT3-RA in the delayed and overall phases [8]. Two large phase III trials involving patients treated with MEC showed higher rate of preventing vomiting with PALO than with ondansetron or dolasetron [23,24]. But in our study, regarding MEC, among patients treated with guideline-consistent dual antiemetics (5HT3-RA + dexamethasone) or triple antiemetics (5HT3-RA + dexamethasone + aprepitant), PALO did not show a higher CR rate when compared with the 1st generation 5HT3-RA. The reason might because in previous clinical trials patients received single agent, while in our clinical setting, most patients received combined antiemetic regimens. Therefore, in terms of combined antiemetic regimens in the real clinical setting, selection between PALO and 1st generation 5HT3-RA needs further veri cation.
Regarding the patients receiving HEC, the addition of aprepitant into dual antiemetics is of great signi cance as triple antiemetics (5HT3-RA + dexamethasone + aprepitant) had higher CR rates in acute and overall phases compared with dual antiemetics (5HT3-RA + dexamethasone). This result was comparable with the study by Navari et al., in which they found addition of aprepitant to a standard regimen of 5HT3-RA and dexamethasone in patients receiving HEC improves the CR rate of acute CINV [9].
Two trials involving patients receiving HEC reported signi cantly higher e cacy in the control of emesis with the addition of aprepitant to ondansetron plus dexamethasone than with ondansetron plus dexamethasone alone [25,26], therefore, the e cacy and effectiveness of aprepitant were both con rmed via previous RCTs and this cross-sectional study in the real clinical setting.
This study identi ed several independent predictors for overall CR in accordance with previous studiesmale, use of aprepitant and no ANV (anticipatory nausea and vomiting). Besides, our results indicated that GCCP, use of aprepitant,two antiemetics (versus single agent) and L/mEC (versus HEC) were the protective factors for overall CR. Differing from the widely accepted view, this study did not identify older age as the protective factor [21]. The reason might be attributed to the median age in our study was 56 years old, which led to an age strata bias.
The main limitation of this study was the data collection was based on patients' recall about the most recent treatment, which led to more missing records compared to the records from patients' daily diary.
Another limitation was the results in this study may not be generalized to suburban areas in China. So above-mentioned weakness could have introduced recall bias and selection bias. Besides, most patients participated in this study were not naïve to chemotherapy, as a result, there might be more confounding factors leading to CINV, such as prior experience of poorly controlled emesis, anxiety about chemotherapy. Despite above limitations, the results in this survey supported previous studies and further illustrated the positive effect of antiemetic guideline on controlling CINV.

Conclusion
This nationwide study reported Chinese CINV status, the prevalence of GCCP and the association between GCCP and CR rate in the real clinical setting for the rst time. The ndings in this study indicated that prevalence of GCCP still have much room for improvement and the incidence of CR in GCCP was higher than that in GICP, thus it might be the best choice for patients to comply with GCCP in the real clinical setting. As such, prescribing antiemetics in compliance with guidelines for all patients receiving chemotherapy is strongly suggested.

Declarations
Ethics approval and consent to participate All methods in this study were carried out in accordance with relevant guidelines and regulations. All experimental protocols were approved by Ethics Committee of Sun Yat-sen University Cancer Center. Informed consent was obtained from all subjects and all subjects are above 18, and the ethical approval was provided in supplement 2.
Manuscript writing: Cong Xue, Lu Li, Qing Xia Final approval of manuscript: All authors