FLOT; To Be Treated as a Highly Emetogenic Regimen or a Moderately Emetogenic One? Comparison of the Emetogenic Potential of FLOT versus FOLFOX and TAC Regimens

Purpose The current study aimed at investigating the ecacy of aprepitant-containing triple antiemetic regimen in FLOT (Fluorouracil+Leucovorin+Oxaliplatin+Docetaxel) recipients as well as the emetogenic potential of FLOT regimen, through comparison of nausea and vomiting rates in a moderately emetogenic chemotherapy, FLOT, and a highly emetogenic chemotherapy recipients. Study regimens, antiemetic guidelines recommend adding olanzapine to aprepitant-containing triple antiemetic regimen besides continuing dexamethasone and olanzapine administration on days 2-4.

Several factors affect the incidence and severity of chemotherapy-induced nausea and vomiting (CINV), being divided into patient-related (including age, sex, history of alcohol consumption, motion sickness, history of pregnancy-associated emesis, anxiety, CINV expectation, concomitant use of opioids, and serotonin speci c reuptake inhibitors-each of the rst three may cause a 10-15% absolute difference in the chance of vomiting with highly emetogenic chemotherapy) and antineoplastic agent-related factors (including intrinsic emetogenicity of the given antineoplastic agent, dose, route of administration and schedule) [3,4].
Since the strongest prognostic factor for emesis is the emetogenic potential of each chemotherapeutic agent [5], Hesketh et al. [6] proposed a ve-level classi cation based on this factor which was determined by the percentage of patients showing acute emesis when chemotherapy was administered in the absence of antiemetic prophylaxis. This emetogenicity schema was modi ed to a four-level classi cation of highly (emetic risk: >90%), moderately (30-90%), low (10-30%) and minimal (<10%) emetogenic agents later in 2004 [7]. Emetic risk determination and appropriate antiemetic prophylaxis selection for antineoplastic combinations are based on the "component antineoplastic agent of greatest emetic risk" [8].
Although Hesketh et al. classi cation system affords important information that builds the very rst steps in proper antiemetic prophylaxis administration and acceptable CINV management in the acute phase, it is accompanied by some limitations; i.e., it underestimates the risk of delayed emesis and, nausea in acute and delayed phases [9] and no row in the table is allocated to each antineoplastic combination for which the net emetogenic potential may be higher than that of the component antineoplastic agent with greatest emetic risk. However, this classi cation system is not rigid and the trials performed concerning CINV made changes leading to improvement of the system over time. The foremost example of the system improvement is Anthracycline and Cyclophosphamide (AC) combination, which was primarily considered to be moderately emetogenic but its substantial risk of nausea and emesis demonstration by trials led to the elevation of its emetic risk level [7].
The limitations of the Hesketh et al. classi cation system also seem to become evident when guidelinedirected antiemetic regimen for managing FLOT-induced nausea and vomiting fails to meet the expectations.
With its demonstrated superiority over ECF/X (Epirubicin, Cisplatin, Fluorouracil/Capecitabine), the FLOT regimen is accepted as the best standard perioperative treatment in locally advanced (≥cT2 and/or cN+ [10]) gastric cancer -the fth most common cancer worldwide with over one million estimated new cases annually [11]-and esophagogastric junction (EGJ) cancer [12]. FLOT is also effective in metastatic gastric and EGJ cancers as a docetaxel-based chemotherapy [10].
According to Hesketh et al. classi cation, FLOT is classi ed as a moderately emetogenic antineoplastic regimen; however, following the administration of the guideline-directed antiemetic regimen (to control nausea and vomiting caused by moderately emetogenic antineoplastic regimen), the percentage of FLOT recipients with nausea and vomiting, especially in the delayed phase, is much higher than what is expected according to previous studies on moderately emetogenic chemotherapies, such as FOLFOX regimen [13]. Hence, it is a prerequisite to determine whether the appropriate antiemetic prophylaxis administration and acceptable CINV management have been done correctly or not. In order to address this issue and investigate the FLOT regimen characteristic from an emetogenic potential aspect, we aimed to compare "no emesis", "complete response (CR)", and "complete protection (CP)" rates in a group of gastric or EGJ cancer patients receiving FLOT, to the rate of mentioned endpoints in two groups of patients: a group of moderately and a group of highly emetogenic chemotherapy recipients when the same three-drug antiemetic combination with aprepitant was administered to all patients as prophylaxis.
To the best of our knowledge, currently there is no study aimed at speci cally assess the e cacy of the aprepitant-containing antiemetic combination in FLOT-receiving patients (the primary objective of this study). Also, no study particularly investigated FLOT regimen characteristic from the emetogenic potential aspect (secondary objective of the current study).
Patients And Methods:

Study Design, Inclusion and Exclusion Criteria:
This was a prospective, observational, comparative study conducted between November 2018 and February 2020, at three centers a liated to Shiraz University of Medical Sciences (SUMS), Shiraz, Iran.
Eligible patients were 18 years and older, with an Eastern Cooperative Oncology Group (ECOG) performance status of 0-2, for whom, locally advanced or metastatic gastric or EGJ cancer (1st group), colorectal cancer (2nd group), and breast cancer (3rd group) was diagnosed by oncologist according to the diagnostic tests and were planned to receive their rst cycle of neoadjuvant or adjuvant FLOT, FOLFOX, and TAC regimens, respectively (chemotherapy-naïve). Molecular targeted therapy was allowed.
A three-drug antiemetic combination including 125 mg-PO aprepitant, 8 mg-IV ondansetron, and 12mg-IV dexamethasone (30 minutes prior to chemotherapy administration) on day 1 and 80 mg-PO aprepitant on days 2 and 3 was administered to all the patients. Exclusion criteria were de ned according to previous studies then, adapted to the research conditions [13,14]. Additional exclusion criteria were receiving radiation therapy to the abdomen or pelvis within 1 week prior to chemotherapy and treatment with CYP3A4 inducers less than four weeks or an inhibitor or substrate of CYP3A4 one week before chemotherapy administration [15].
This study is conducted in compliance with the principles of the Declaration of Helsinki and the national norms and standards for conducting medical research in Iran .All patients provided written informed consent to participate in the study.

E cacy Endpoints:
The primary e cacy endpoint was the proportion of patients with "CR" (no emesis and no rescue medication use) during 120 hours after initiation of chemotherapy (overall phase) in cycle 1.
Key secondary e cacy endpoints included "CR" during the acute and delayed phases as well as "no emesis" and "CP" (no emesis, no rescue medication use, no more than mild nausea) during acute, delayed, and overall phases in the cycle 1. As exploring rescue medication use, at the end of the study, revealed that all rescue medication consumers had vomiting and/or retching and/or great/severe nausea, "CR" was also de ned as no emesis and less than great/severe nausea.
Time-to-antiemetic treatment failure (rescue medication use which was due to emesis or great/severe nausea) and time-to-rst emetic episode were other endpoints of this study.
"No signi cant nausea" (none or mild nausea) rate in acute, delayed, and overall phases is also reported to evaluate the e cacy of the aprepitant-containing antiemetic combination in FLOT-receiving patients.

Assessment:
Patients were assessed prior to chemotherapy and then every 24 hours from the start of chemotherapy till the end of day 5. Any nausea, vomiting, and retching occurring within 24 hours after the initiation of chemotherapy was de ned as acute, while nausea, vomiting, and retching occurring between 24 and 120 hours after the start of the chemotherapy was de ned as delayed.
An emetic episode was de ned as any episode of vomiting or retching (non-productive vomiting) or combined vomiting/retching. An interval of at least one minute between two emetic episodes was required to characterize discrete emetic episodes.
Culturally and linguistically adapted, translated-into-Persian version of Rhodes Index of Nausea, Vomiting and Retching (Rhodes INVR) [16] was applied to investigate the frequency and distress (none, mild, moderate, severe, great) of nausea, vomiting, and retching in the round on patient prior to chemotherapy and at the end of the acute phase and also, on daily phone conversations in the delayed phase.
All the patients were trained to take ondansetron in acute and metoclopramide in the delayed phase if an emetic episode and/or great to severe nausea occurred. Rescue medication use was then reported to investigators.
Aprepitant-containing antiemetic regimen e cacy in FLOT recipients (the primary objective of the study) was reported as frequency and proportion of patients achieving e cacy endpoints.
In order to present patient characteristics, the only continuous variable, i.e. age, was summarized as mean (Std. Deviation) and categorical variables as frequencies and proportions. To explore characteristics distribution homogeneity among three different antineoplastic regimen groups, one-way ANOVA and chi-square (or Fisher's exact) tests were applied for continuous (age) and categorical variables, respectively.
Inter-group comparison of the number of emetic episodes in acute, delayed and overall phases, separately, was performed using Kruskal-Wallis test and the different group was determined by pairwise comparisons. The Bonferroni correction was used to adjust for multiple comparisons.
Logistic regression was applied not only to adjust for covariates (including age, sex, BMI (Body Mass Index), concomitant opioid consumption, concomitant tobacco/cigarettes consumption, motion sickness, CINV expectation, alcohol consumption, neoadjuvant/adjuvant chemotherapy, stage of cancer and history of pregnancy-associated emesis) in assessing antineoplastic regimen effect on the risk of acute, delayed and overall CINV, being expressed as odds ratios (OR) with 95% con dence intervals but also, to explore the risk factors for CINV. Hence, univariable analysis was performed primarily, then multivariable logistic regression analysis (Wald's test) by the backward method was performed on items with a signi cance level of less than 20%.
Cox proportional hazards regression analysis was performed to estimate hazards ratios (HR) and 95% con dence intervals for the time-to-antiemetic regimen failure (vomiting, retching, or severe nausea occurrence) and time-to-rst emetic episode, concerning the different antineoplastic regimens. Results:

Patients:
Among a total of 182 chemotherapy-naïve patients who were enrolled in this study, 165 patients (63 patients in the TAC group, 52 patients in FLOT group and 50 patients in FOLFOX group) met the eligibility criteria and were included in the analysis. In the TAC group, migraine was the leading cause of exclusion while in the FLOT group, tumor site causing obstruction resulted in patients' ineligibility.
The mean age (Std. Deviation) of the patients in the FOLFOX, FLOT, and TAC groups were 57.34 (11.621), 55.60 (12.693), and 42.57 (7.219), respectively. The other baseline characteristics of the patients are described in Table 1. The Low rate of positive alcohol consumption, concomitant use of opioids, and motion sickness may lead to the lack of power in testing the relationship between these factors and CINV occurrence.

E cacy of Aprepitant-Containing Triple Antiemetic Regimen in FLOT Recipients:
The primary e cacy endpoint, "CR"-overall phase, was achieved in 61.5% (32/52) of FLOT recipients. The percentage of the FLOT-receiving patients achieving other e cacy endpoints in acute, delayed, and overall phases are shown in Fig. 1.

FLOT Regimen Characteristic from Emetogenic Potential Aspect:
In order to explore the emetogenic potential of the FLOT regimen (secondary objective of the current study), we compared "CR", "no emesis" and "CP" rates in acute, delayed, and overall phases among FLOT and each of FOLFOX and TAC groups. The frequency and proportion of the patients achieving each endpoint are shown in Table 2. Comparison among FLOT and each of FOLFOX and TAC groups, concerning endpoint-achieving rates, was performed using logistic regression to adjust for covariates. As shown in Table S 1 . (Online Resource 1, which presents the odds of achieving each endpoint; adjusted for covariates), the odds ratios indicate the signi cant difference among FLOT and FOLFOX regimens and non-signi cant difference among FLOT and TAC in relation to the rates of "CR", "no emesis" and "CP" in the delayed and overall phases, also "CP" in the acute phase. The difference is signi cant among FLOT and FOLFOX, also, FLOT and TAC concerning "CR" and "no emesis", in the acute phase. To explain in detail: The odds of the FOLFOX recipients achieving "CR" in the overall phase is 3.145 times as high as the odds of FLOT recipients (95%C. In comparison to FLOT recipients, the odds of achieving "CP" in the overall phase is 3.784 and 2.354 times greater for FOLFOX (C.I.: 1.459-9.816, p=0.006) and TAC (C.I.: 0.782-7.088, p=0.128) patients, respectively. The FOLFOX recipients' odds of achieving "CP" in the delayed phase is approximately three and a half fold greater (OR=3.428, C.I.: 1.355-8.669, p=0.009) than the FLOT recipients' odds and TAC recipients' odds is more than one and a half fold higher (OR=1.672, C.I.: 0.674-4.149, p=0.267) than FLOT recipients' odds.
While antineoplastic regimen is the only signi cantly associated factor (multivariable logistic regression) with "no emesis" in the delayed phase, antineoplastic regimen and sex are both the statistically signi cant associated ones with "no emesis" in acute and overall phases, also "CR" and "CP" in the acute phase. Antineoplastic regimen and CINV expectation are statistically signi cant predictive factors for "CR" not only in the delayed phase but also in the overall phase.
Antineoplastic regimen, age, and CINV expectation are signi cantly associated with "CP" in the delayed phase, and antineoplastic regimen, age, and sex are the signi cant predictive factors for "CP" in the overall phase.
As mentioned earlier, the history of pregnancy-associated emesis was explored only in 55.15% (91/165) of the patients who had pregnancy experience and it was statistically signi cant in the univariable analysis only for "CR" in delayed and overall phases. Hence, multivariable logistic regression (Backward-Wald's test) was performed again on these endpoints including all previously recognized signi cant predictive factors (p-value<0.2 in univariable analysis except sex) and history of pregnancy-associated emesis. The outcome clari es that the FOLFOX recipients' odds of achieving "CR" in the delayed phase is more than seven and a half fold greater (OR=7.723, C. 0.138-0.824, p=0.017). Also, as reported previously [19], most of the patients who showed emesis, started vomiting and/or retching within the rst three days.
Kruskal-Wallis test was performed to compare the number of emetic episodes among groups and assisted in achieving the secondary objective of the current study. As shown in Table 3, the difference between the number of emetic episodes is statistically signi cant in the delayed (p=0.012) and overall (p=0.011) phases and pairwise comparison revealed that FOLFOX is the different regimen. In the acute phase, the number of emetic episodes is not signi cantly different among groups (p=0.161). Discussion: Over the last few decades, a better understanding of the complicated multifactorial pathophysiology of CINV, which involves interaction among different neurotransmitters and receptors in the central nervous system and gastrointestinal tract, caused important advances in CINV management, however, CINV prevention often remains suboptimal particularly for chemotherapy-induced nausea in delayed phase [20,21]. Therefore, there is an ongoing effort involved in paving the way to better CINV management, from improving emetic risk determination to nding new targets to design antiemetic agents.
A novel scenario in CINV management is FLOT-induced nausea and vomiting. Although the FLOT regimen is considered to be moderately emetogenic, guideline-directed antiemetic combination to manage moderately emetogenic chemotherapy does not meet the expectations when administered to FLOT recipients. This can be considered as the main reason to perform the present study aiming at investigating the e cacy of aprepitant-containing triple antiemetic regimen in FLOT recipients and FLOT characteristic from the emetogenic potential aspect by comparing FLOT recipients to moderately (FOLFOX) and highly (TAC) emetogenic chemotherapy recipients. FOLFOX and FLOT are almost alike (both are platinum-based agents which are characterized by their ability to induce delayed nausea and vomiting) except for the rate of 5-FU administration (2600mg/m 2 -administered in 20 hours in FLOT and 2400mg/m 2 -administered in 40 hours in FOLFOX) and the presence of docetaxel in FLOT combination so it seems to be the best choice among moderately emetogenic regimens. TAC, being at the top of the highly emetogenic agents list in the emetogenic potential classi cation table, is the best choice among highly emetogenic regimens. The proportion of the patients with "no emesis", "complete response", "complete protection" and "no signi cant nausea" in acute, delayed, and overall phases in the FLOT-receiving group (N=52) Figure 2 Plot for Cox proportional hazards model compares the proportion of patients with "CR" between different antineoplastic regimens, at any point of time in the overall phase, adjusted for covariates