Chemotherapy-induced nausea and vomiting is less controlled at delayed phase in patients with esophageal cancer: a prospective registration study by the CINV Study Group of Japan

SUMMARY

Chemotherapy is an indispensable therapeutic approach for esophageal cancer. Although chemotherapy-induced nausea and vomiting (CINV) is one of the most crucial adverse events, the current state of CINV in patients with esophageal cancer remains unclear. This multicenter prospective observational study analyzed data for 192 patents with esophageal cancer who underwent moderately emetogenic chemotherapy (MEC) or highly emetogenic chemotherapy (HEC). The patients recorded their CINV incidence and severity daily for 7 days after receiving chemotherapy, using visual analog scales (VAS). Of the 192 patients, 181 received HEC including cisplatin, and 11 patients received MEC including nedaplatin. Approximately 81% of HEC and 82% of MEC patients received antiemetic therapy in compliance with guidelines. Although CINV was controlled relatively well in the early phase (days 1–4), it was not fully controlled in late phase (days 5–7) for both the HEC and MEC groups. Female sex was a major risk factor for delayed vomiting (P=0.034). Multivariate logistic regression analysis for VAS revealed that motion sickness, age, and use of other antiemetics were risk factors for delayed nausea. Adherence to antiemetic guidelines effectively controls vomiting but is less effective against delayed CINV in both HEC and MEC patients. Identification of individual risk factors, such as female sex, will help develop personalized treatments for CINV. In the clinical setting for esophageal cancer, regimens that include nedaplatin might need to be treated as HEC.

INTRODUCTION

Esophageal cancer is the sixth most common cause of cancer-related death, and the eighth most commonly diagnosed cancer worldwide.¹ In the East, including Japan, its predominant histological type is squamous cell carcinoma.² Although esophageal squamous cell carcinomas (ESCCs) are generally treated by surgical resection, its high rates of systemic and local tumor recurrence have led to investigations into multimodality therapies that combine surgery with radiotherapy, chemotherapy, and chemoradiotherapy. Recently, the Japan Clinical Oncology Group study (JCOG9907) demonstrated that preoperative chemotherapy with cisplatin (CDDP) plus 5-fluorouracil (5-FU) followed by surgery improves the overall survival of patients with resectable thoracic ESCC.³ As then, preoperative chemotherapy followed by radical esophagectomy has been accepted as the standard therapeutic approach for resectable cStage II/III ESCC in Japan. Similarly, the regimen with CDDP and 5-FU has been accepted as the first-line treatment for metastatic and unresectable ESCCs in Japan.^4–6 Chemotherapy is an indispensable therapeutic approach for both resectable and unresectable ESCCs.

Chemotherapy-induced nausea and vomiting (CINV) is a critical adverse event among patients with esophageal cancer who receive chemotherapy. CINV leads to reduced dietary intake, which is associated with poor nutrition status and can, in turn, postpone the next administration of chemotherapy, or may even cause patients to refuse further treatment. Poor control of CINV reportedly reduces the relative dose intensity of chemotherapeutic agents, leading to poor treatment outcome and shorter patient survival.⁷ A combination of a 5HT₃ receptor antagonist (5HT₃RA) and dexamethasone is recommended for moderately emetogenic chemotherapy (MEC), and a combination of three antiemetics consisting of these two drugs and aprepitant is recommended for highly emetogenic chemotherapy (HEC).^8,9 In Japan, the guidelines proposed by the Japanese Society of Clinical Oncology (JSCO) have gradually led oncologists to administer prophylactic antiemetic drugs,¹⁰ instead of administering them therapeutically after the development of CINV. Nonetheless, the current state of CINV in esophageal cancer patients remains unclear.

We, therefore, conducted a Japanese multicenter surveillance study to clarify the incidence of CINV and the use of antiemetics, in accordance with the

JSCO guideline in esophageal cancer patients. Second, as this study investigated the incidence of CINV for up to 7 days after chemotherapy, we could evaluate not only early CINV but also delayed CINV. Third, the risk factors for CINV were also explored in this study.

MATERIALS AND METHODS

Study design

A nationwide, multicenter prospective cohort study to survey CINV in patients with esophageal carcinoma was conducted by the CINV Study Group of Japan.¹¹ We selected university hospitals, cancer centers, and cancer treatment hospitals that were certified by the Ministry of Health, Labour and Welfare of Japan and asked them to participate in this study. One hundred and eight institutions throughout Japan from the southern district of Japan, Kyushu to the northern part of Japan, Hokkaido understood the study purpose and agreed to participate. The study was approved by the institutional review board in each hospital and written informed consent was obtained from all patients.

Enrollment of patients

We enrolled patients who were scheduled to receive cancer chemotherapy with moderate or high emetogenicity for the first time. Classification of the emetogenic risk of anticancer drugs was based on the ‘Guidelines for Appropriate Use of Antiemetic Drugs, Version 1’ published by JSCO,¹⁰ which were themselves based on the NCCN Clinical Practice Guidelines in Oncology -Antiemetics- ver. 4, 2009.¹²

The registration forms recorded background information including patients' initials, date of birth, sex, hospital patient number, type of cancer, disease stage, and other eligibility criteria. The forms were filled out by the investigators and faxed to the secretariat of the central office located at Fukuoka University before commencing cancer chemotherapy. The original registration forms were preserved in the clinical records. To investigate estimation of CINV by the medical staff (i.e. doctors and nurses), staff members were also requested to fill out questionnaires about the severity of digestive symptoms in the acute and late phases of CINV on the same registration form, if they thought their patients would experience CINV. Acute and delayed CINV were defined as nausea and vomiting that developed within or after 24 hours after the start of chemotherapy, respectively. The secretariat confirmed that each patient matched the inclusion criteria but not the exclusion criteria based on the registration form, and then issued the patient with a registration number.

Collection of patient diaries and case reports

Patients were provided with 7-day diaries for recording CINV prior to commencement of cancer chemotherapy. They were asked to record their digestive symptoms, i.e. development and severity of nausea, frequency of vomiting, amount of food intake, number of salvage treatments, and confirmation of hospitalization or outpatient visits. Patients were required to fill in their diaries every day for 7 days from commencement of anticancer MEC or HEC. Severity of nausea and oral intake were checked with visual analog scales (VAS) using linear scales augmented by facial scales. Occurrence of nausea was defined when a VAS was measured 25 mm or longer.

The investigators and/or their colleagues recorded background patient information, including patient initials, sex, hospital patient number, date of birth, treatment history (history of radiotherapy, use of an anticancer drug with mild or low emetogenicity, use of an anxiolytic drug before administration of the anticancer drug, and use of opioids), alcohol intake history, risk factors for CINV (history of motion sickness or vomiting related to pregnancy), performance status, clinical test results (hemoglobin and albumin levels), cancer chemotherapy regimen (type and dose of drug and administration period), and details of antiemetic therapy and salvage treatment for CINV in the case report. Patients were requested to fill in their diaries and send them to the central office after putting them in the provided return envelopes. The case reports were also sent to the secretariat by the investigators, using the return envelopes.

Treatment

The Japanese guidelines for CINV published in 2010 recommend the use of either two antiemetics consisting of a 5HT₃RA and dexamethasone for use with MEC, or three antiemetics, consisting of the above antiemetics plus aprepitant for use with HEC.¹⁰ Aprepitant can also be added for use with MEC regimens that contain carboplatin, irinotecan, ifosfamide, or methotrexate. The dose and duration of dexamethasone treatment differ between HEC and MEC regimens. In patients who received HEC, 9.9 mg of dexamethasone was given intravenously on Day 1 followed by 8 mg orally for a maximum of 4 consecutive days; whereas in patients who received MEC, half the dexamethasone dose was given intravenously on Day 1, followed by 4 mg orally for a maximum of 3 days.

Data analysis

Patient demographics and survey responses were summarized using descriptive statistics. The numbers of risk factors for CINV in the collected data sets were analyzed by multivariate logistic regression analysis (Wald's test). Cohen's kappa coefficients were calculated to estimate the concordance between actual and predicted CINV by medical staff. All reported P-values correspond to two-sided tests; P < 0.05 was considered to be significant. Analyses were performed with SAS for Windows release 9.3 (SAS Institute, Cary, NC, USA).

RESULTS

Background characteristics

In this study, we analyzed 192 patients who were affected with esophageal cancer. Their baseline characteristics (e.g. age, sex, tumor stage, and Performance status [PS]) are shown in Table 1. Most patients (85%) were male. Almost all patients had a good PS (0 or 1 on the ECOG scale; 99%). Chemotherapy regimens are shown in Table 2. Of the 181 patients who received HEC, more than 50% (n=102) received CDDP + 5-FU, which is accepted in Japan as the standard neoadjuvant regimen for clinical Stage II/III ESCCs, and the first-line regimen for metastatic/unresectable ESCCs.^3,4 In the HEC group, 40% (n=72) received triplet chemotherapy, using docetaxel combined with CDDP + 5-FU, which reportedly has strong antitumor activity for esophageal cancer and is a promising candidate for the next standard ESCC chemotherapy regimen. All 11 MEC patients received nedaplatin-based chemotherapy.

Table 1

Patient characteristics

			N	%
Age: mean (range)		Total	66 (40–84)
Sex		Male	164	85
		Female	28	15
		Total	192	100
Stage	Primary	Stage I	9	4.7
		Stage II	48	25
		Stage III	66	34
		Stage IV	60	31
	Recurrence		9	4.7
PS (ECOG)		0	147	77
		1	42	22
		2	3	1.6

Table 2

Chemotherapy regimen and antiemetic use

HEC (n=181)		MEC (n=11)
Regimen	Number (%)	Regimen	Number (%)
CDDP+5-FU	102 (56%)	docetaxel+ nedaplatin+ 5-FU	5 (46%)
CDDP+ 5-FU+ docetaxel	72 (40%)	docetaxel+ nedaplatin+ S-1	3 (27%)
CDDP+ 5-FU+ adriamycin	5 (2.8%)	Nedaplatin+ 5-FU	2 (18%)
Others	2 (1.2%)	nedaplatin	1 (9.1%)
No of antiemetics		No of antiemetics
3 antiemetics	147 (81%)	3 antiemetics	7 (64%)
2 antiemetics	24 (13%)	2 antiemetics	2 (18%)
Others	10 (5.6%)	Others	2 (18%)

Use of antiemetics

Antiemetic regimens were HEC group: 147 patients (81%) received three antiemetics, 24 (13%) received two, with ∼81% receiving guideline-compliant therapy; MEC group: 7 patients (64%) received three antiemetics, 2 (18%) received two, with ∼82% receiving guideline-compliant therapy.¹⁰ Breakthrough CINV was treated mostly by adding an antiemetic agent with a different mechanism of action from the prophylactic drugs, or changing to a different 5HT₃RA.

Control of CINV

Incidence patterns for CINV are shown in Figure 1A. Acute CINV was well controlled (acute nausea and vomiting with HEC, 4.4% and 3.9%; with MEC, 0% and 0%, respectively). Delayed nausea was common (41% with HEC and 46% with MEC), but the incidence of delayed vomiting was relatively low (13% with HEC and 18% with MEC).

Fig. 1

(A) Occurrence of acute/delayed nausea and vomiting in HEC- and MEC-induced CINV. (B) VAS for nausea severity from Days 1–7.

Analysis of risk factors

Results of univariate and multivariate logistic regression analysis for delayed vomiting are shown in Table 3. Female sex was an independent risk factor for delayed vomiting in patients with esophageal cancer (odds ratio: 2.97, 95% confidence interval: 1.09–8.14, P=0.034). Previously known risk factors (e.g. history of motion sickness, drinking habits) were not identified as risk factors for delayed vomiting. No factors (such as sex, motion sickness, drinking habit, or age) were associated with the acute or delayed occurrence of nausea (data not shown).

Table 3

Prognostic factors for delayed vomiting

	Univariate analysis		Multivariate analysis
Factors	OR (95%CI)	P value	OR (95%CI)	P value
Sex: Female versus Male	2.70 (1.01–7.25)	0.048	2.97 (1.09–8.14)	0.034
Motion sickness: Yes versus No	4.42 (0.99–19.8)	0.052
Drinking habit: Yes versus No	1.33 (0.53–3.38)	0.55
Age	0.96 (0.91–1.01)	0.090	0.95 (0.91–1.00)	0.066
3 antiemetics versus 2 antiemetics	0.77 (0.24–7.03)	0.67
Other antiemetics versus 2 antiemetics	1.10 (0.17–7.03)	0.92
HEC versus MEC	0.66 (0.13–3.22)	0.60

CINV incidence and severity pattern

Severity of nausea during days 1–7 after chemotherapy was measured by VAS (Fig. 1B). As the incidence of vomiting was too low for a visible difference in the CINV pattern, we analyzed only the incidence of chemotherapy-induced nausea. In HEC patients, nausea was controlled relatively well in the early phase (days 1–4), but not in the late phase (days 5–7). Importantly, a similar tendency was seen in MEC patients. Multivariate logistic regression analysis for VAS showed that risk factors for measurable nausea included motion sickness (P=0.021), age (P=0.0061), and use of other antiemetics (P=0.042). Based on these results, we conducted a linear mixed model for VAS, using motion sickness, age, and use of other antiemetics (Fig. 2).

Fig. 2

Linear mixed model for Visual Analog Scales (VAS) for nausea, using motion sickness, age, and other antiemetic use. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Estimation of CINV by the medical staff

Estimation of CINV by medical staff is shown in Table 4. Staff generally overestimated CINV development, especially nausea. Medical staff estimated that 89 patients (46%) would experience acute nausea, but only 8 (4.2%) actually developed nausea. Similarly, in terms of delayed nausea, medical staff estimated that 178 patients (93%) would experience delayed nausea, but 79 (41%) actually developed delayed nausea. Similar results were observed for acute/delayed vomiting.

Table 4

Estimation of CINV by medical staff compared with actual CINV

		Estimation by medical staff					Estimation by medical staff
		No CINV	CINV	Total			No CINV	CINV	Total
Acute nausea	Absent	98	86	184	Delayed nausea	Absent	7	106	113
	Present	5	3	8		Present	7	72	79
	Total	103	89	192		Total	14	178	192
	K value	−0.0158				K value	−0.0242
Acute vomiting	Absent	148	37	185	Delayed vomiting	Absent	142	25	167
	Present	5	2	7		Present	18	7	25
	Total	153	39	192		Total	160	32	192
	K value	0.0268				K value	0.1164

DISCUSSION

This is the first prospective nationwide CINV survey of patients with esophageal cancer. We found that more than 80% of esophageal cancer patients received antiemetic therapy in compliance with the guidelines in Japan. Second, CINV was controlled relatively well in the early phase (days 1–4), but not in the late phase (days 5–7), in both the HEC and MEC groups. Third, female sex was an independent risk factor for delayed vomiting. Fourth, multivariate logistic regression analysis of VAS data showed motion sickness, age, and use of other antiemetics were risk factors for delayed nausea.

As the regimen that included CDDP is usually selected for esophageal cancer chemotherapy in Japan,^13–15 almost all regimens were HEC (94%) in this study. All guidelines, including JSCO guidelines, recommended that patients on HEC receive three antiemetics, including a 5HT3RA, NK1RA, and dexamethasone. In our study, 147 of 181 patients (81%) on HEC received three antiemetic drugs within the guidelines. In the MEC group, the rate of compliance with CINV guidelines was 82%. In the Pan European Emesis Registry (PEER) study (which was aimed at several types of malignancies), adherence to guidelines for CINV prophylaxis in cycle 1 was 55% and 46% for the acute and delayed phases, respectively.¹⁶ According to the Oncology Analyzer database of IMS Health, aprepitant was used only in 25% of HEC patients in European countries (i.e. France, Germany, Italy, Spain, and UK), which implies that 75% used either dexamethasone alone or dexamethasone plus a 5HT₃RA. In addition, the INSPIRE study showed antiemetic guideline adherence for patients who receive HEC to be 91% for the acute phase and 29% for the delayed phase in US community oncology practices.¹⁷ These results may suggest that antiemetic guideline consistency varies among countries and oncologists.

Although most studies have only investigated CINV incidence for 5 days after chemotherapy, this study asked patients to record their digestive symptoms for up to 7 days after chemotherapy, which allowed us to find that VAS scores for both HEC and MEC were on a rising trend even at day 7. Although higher guideline compliance might decrease CINV incidence, especially in early phase, late-phase CINV remains a huge problem. Future studies of CINV might therefore collect information for longer than 7 days. Additionally, the mechanism of delayed CINV should be studied in greater depth to find better treatments against delayed CINV.

Interestingly, we found that MEC for esophageal cancer induces CINV in a manner similar to that of HEC. The guidelines suggest that MEC regimens that contain carboplatin, irinotecan, ifosfamide, and methotrexate could be candidates for adding aprepitant to dexamethasone and a 5HT₃RA. As combination chemotherapy that uses nedaplatin and 5-FU is reported to be safe and effective for treating esophageal cancer, nedaplatin is accepted as a key chemotherapeutic drug.^18,19 As all MEC patients in this study received nedaplatin-based chemotherapy, we may need to treat regimens with nedaplatin as HEC in clinical settings for patients with esophageal cancer. These findings needs be confirmed by independent studies with a much larger sample size or well-designed prospective studies.

Identification of individual risk factors is important in the development of personalized treatments for CINV. We found that female sex is a major risk factor for delayed vomiting. In addition, motion sickness, age, and use of other antiemetics are risk factors for VAS scores that indicate delayed nausea. We plan to conduct an exploratory analysis of the results of the present study to investigate more optimal strategies for antiemetic-based control of CINV, which would consider the patient's risk factors, as well as the emetogenicity of the chemotherapy regimen.

The finding on CINV estimation by health professionals is intriguing. Previous studies reported that medical staff underestimated CINV development.^20,21 However, this current study reached an opposite conclusion: Japanese investigators overestimated the likely incidence of CINV, which indicates that Japanese medical personnel fully understand the importance of CINV. Overestimation might reflect good control of CINV, especially in relation to the low incidence of vomiting associated with the high rate of use of three antiemetics in patients with HEC regimens, in accordance with the CINV guidelines.

In conclusion, compliance with CINV guidelines in esophageal cancer chemotherapy was relatively good and led to good control of chemotherapy-induced vomiting in Japan. However, the incidence of delayed nausea remained high in both HEC and MEC groups. Female sex was a major risk factor for delayed nausea in either regimen. Further studies are needed to develop more individualized regimens based on each patient's risk.

Acknowledgments

We also would like to thank all the patients and investigators who participated in this study.

SUPPORTING INFORMATION

Additional Supporting Information may be found in the online version of this article at the publisher's web-site:

Fig. S1 •••.