Delayed and long-delayed chemotherapy-induced nausea and vomiting in moderately emetogenic chemotherapy: a scoping review of reporting gaps

Abstract

Background

Chemotherapy-induced nausea and vomiting (CINV) remains a clinically significant adverse event in patients receiving moderately emetogenic chemotherapy (MEC). Although the delayed phase is conventionally defined as 24-120 h after chemotherapy, symptoms may persist beyond this timeframe. However, the extent to which long-delayed CINV (>120 h) has been evaluated in prospective MEC studies remains unclear.

Patients and methods

This study aimed to systematically map the definitions and reporting of delayed and long-delayed CINV in prospective MEC trials, with an exploratory quantitative synthesis of reported CR rates across assessment periods. PubMed/Cochrane Library/Medical Online/Ichushi-Web was used to search for studies published between January 2000 and December 2025. Eligible studies reported CR rates and clearly defined delayed-phase outcomes. A meta-analysis of proportions was carried out to estimate the pooled CR rates in the acute, delayed, and long-delayed phases.

Results

Twenty-six studies comprising 15 phase III and 11 phase II trials were included. The delayed phase was consistently defined as 24-120 h, with 80.8% of studies restricting outcome assessment to this timeframe. Only a limited number of studies evaluated CINV beyond 120 h. In these studies, the pooled CR rate declined from 71% during the conventional delayed phase to 62% during the long-delayed phase. Exploratory subgroup analyses suggested higher pooled CR rates with neurokinin-1 receptor antagonist–containing regimens and in studies including older patients, whereas lower rates were reported in studies including higher proportions of female patients.

Conclusions

In prospective trials, delayed CINV is almost uniformly evaluated within 120 h. CR rates decline in the long-delayed phase, suggesting that conventional endpoints may overestimate antiemetic efficacy. Extending assessment beyond 120 h may provide a more comprehensive characterization of CINV and should be considered in future prospective trials, particularly for newer agents associated with prolonged emetic risk.

Highlights

• •Delayed CINV is almost uniformly evaluated within 120 h despite evidence that symptoms may persist beyond this period.
• •CR rates decline in the long-delayed phase, suggesting that conventional endpoints may overestimate antiemetic efficacy.
• •Extending the evaluation beyond 120 h is feasible and may be necessary to better characterize CINV.

Introduction

Chemotherapy-induced nausea and vomiting (CINV) substantially impairs patients’ quality of life1, 2, 3 and remains one of the most distressing adverse events associated with cancer treatment.4, 5, 6, 7 Beyond symptom burden, poorly controlled CINV may reduce adherence to anticancer therapy and compromise treatment continuity.^8,9 Therefore, effective antiemetic prophylaxis is essential to maintain dose intensity and optimize clinical outcomes.

For patients receiving moderately emetogenic chemotherapy (MEC), current international guidelines recommend a doublet antiemetic regimen consisting of 5-hydroxytryptamine-3 receptor antagonists (5-HT₃RAs) and dexamethasone (DEX).10, 11, 12, 13 However, control with this regimen is often insufficient in oxaliplatin- or carboplatin-based chemotherapy and escalation to a triplet regimen, including a neurokinin-1 receptor antagonist (NK1RA), or the addition of olanzapine may be considered appropriate. Advances in antiemetic therapy have markedly improved the control of acute CINV (0-24 h), particularly with the introduction of second-generation 5-HT₃RAs, such as palonosetron (PALO) and NK1RAs. The combination of NK1RAs and PALO has demonstrated superior efficacy in preventing delayed-phase symptoms (24-120 h) in several trials.14, 15, 16

However, delayed CINV may remain under-recognized, especially in outpatient settings and regimens not containing cisplatin.¹⁷ Consequently, delayed-phase control rates may be overestimated in clinical practice. Even with NK1RA-containing regimens, 10%-20% of patients continue to experience delayed CINV,14, 15, 16^,18,19 indicating persistent gaps in prevention. Moreover, symptoms extending beyond 120 h (long-delayed phase) may reduce adherence or contribute to treatment discontinuation, potentially worsening clinical outcomes.²⁰ Recent individual patient–level analyses have further suggested that symptoms may persist beyond the conventional 120-h boundary.²¹ In a large meta-analysis of patients receiving highly emetogenic chemotherapy (HEC) and MEC, 24% of patients receiving MEC experienced long-delayed nausea (>120 h) and 6% experienced long-delayed vomiting.²² Although the addition of an NK1RA to MEC has been debated and some analyses have questioned its incremental benefit, clinical trials, such as the SENRI study (a multicenter, randomized phase 3 trial evaluating aprepitant in patients with colorectal cancer receiving oxaliplatin-based chemotherapy),¹⁴ have shown that adding an NK1RA to PALO and DEX improved delayed-phase complete response (CR; no vomiting and no rescue medication) and complete protection (no vomiting, no rescue medication use, and no moderate or worsened nausea) over 120 h in oxaliplatin-based chemotherapy.

Historically, the conventional delayed phase of CINV was established on the basis of the biphasic pattern of nausea and vomiting observed with cisplatin-based chemotherapy, in which delayed symptoms generally resolved within ∼120 h after treatment. This temporal framework was adopted by the Multinational Association of Supportive Care in Cancer (MASCC) Perugia Consensus Conference in 1998²³ and subsequently accepted by regulatory authorities as the standard assessment window for delayed CINV.

Importantly, nearly all clinical trials evaluating MEC have defined the delayed phase as occurring within 24-120 h after chemotherapy. Although this 120-h cut-off has become the conventional boundary for delayed CINV assessment, symptoms may persist beyond this timeframe. However, the extent to which the extended (>120 h) outcomes have been systematically evaluated in MEC remains unclear.

In HEC, several clinical trials evaluating long-acting NK1RAs have extended the observation period for CINV outcomes beyond the conventional 120-h window, in some cases up to 168 h. These studies suggest that clinically relevant symptoms may persist beyond 120 h and that extended evaluation is feasible within prospective clinical trial frameworks.

However, it remains unclear whether similar temporal evaluations have been systematically carried out in MEC, where the CINV risk is clinically meaningful; however, trial designs have historically focused on the conventional delayed phase.

Given these considerations, we conducted a scoping review to map how the delayed (24-120 h) and long-delayed (>120 h) phases have been defined and reported in prospective clinical studies on MEC.

Methods

Literature search strategy

This scoping review was conducted to map the definitions and reporting of the delayed (24-120 h) and long-delayed (>120 h) phases of CINV in MEC. An initial structured search of PubMed, Cochrane Library, Medical Online, and Ichushi-Web was carried out to identify prospective clinical trials that evaluated antiemetic prophylaxis in adult patients receiving MEC regimens. Search terms included combinations of the following: ‘moderately emetogenic chemotherapy’, ‘carboplatin’, ‘oxaliplatin’, ‘irinotecan’, ‘chemotherapy-induced nausea and vomiting’,⁵ ‘delayed phase’, ‘complete response’, and ‘specific antiemetics’ [e.g. PALO, aprepitant, fosaprepitant, rolapitant, netupitant (NTP)–PALO fixed combination [NEPA], olanzapine, and mirtazapine]. The timeframe included studies published between January 2000 and December 2025. To ensure the inclusion of contemporary evidence and guideline-aligned antiemetic strategies, the search was subsequently updated on the basis of current antiemetic guideline recommendations, including the National Comprehensive Cancer Network, American Society of Clinical Oncology (ASCO), MASCC/European Society for Medical Oncology (ESMO), and Japanese Society of Clinical Oncology Antiemesis Guidelines.10, 11, 12^,24 Two reviewers (YI and SK) independently screened the titles and abstracts of all the identified records for eligibility. Full-text articles were independently assessed by both reviewers on the basis of the predefined inclusion and exclusion criteria. Any disagreements were resolved through discussion and, when necessary, a third reviewer was consulted to reach a consensus. Although this study was designed as a scoping review to map the definitions and reporting of delayed and long-delayed CINV, a quantitative synthesis was carried out to provide a descriptive summary of the CR rates across different periods. This approach was intended to complement the primary objective of evidence mapping by illustrating temporal trends in antiemetic efficacy rather than formally comparing treatment effects or establishing definitive efficacy estimates. Therefore, a meta-analysis of these proportions was carried out as an exploratory component within the scoping review framework.

Eligibility criteria

Studies were included if they (i) were prospective clinical trials (phase II or III), (ii) included adult patients receiving MEC regimens, (iii) reported CR rates, and (iv) clearly defined a delayed-phase time window. Studies were excluded if they (i) evaluated only the acute phase (0-24 h), (ii) did not separate delayed-phase outcomes, (iii) were retrospective analyses, and (iv) had duplicate publications or secondary subgroup analyses of previously included trials, unless they provided novel time-window data.

Given that emetogenic risk classifications have evolved over time and may differ across contemporary guidelines, eligibility was determined according to the emetogenic classification applied at the time each study was designed and conducted. Moreover, we acknowledge that some regimens or agents included in this review may be classified or managed differently in current guidelines. In particular, anthracycline–cyclophosphamide (AC) regimens are now generally classified as highly emetogenic. Furthermore, the classification and recommended antiemetic management of carboplatin-based chemotherapy and trastuzumab deruxtecan are not uniform across contemporary guidelines.

Data extraction and classification

For each eligible study, the following data were extracted: (i) study design [phase III randomized controlled trial (RCT), phase II RCT, or single-arm], (ii) chemotherapy regimen, (iii) antiemetic strategy, (iv) definition of the delayed phase, (v) CR rates during 24-120 h, (vi) evaluation beyond 120 h (if carried out), and (vii) CR rates during the long-delayed phase. Studies were categorized according to antiemetic strategy and evaluation time window (0-24 h, 24-120 h, and <120 h).

Data synthesis

Data were descriptively synthesized given the heterogeneity in study designs, patient populations, and outcome definitions. Study characteristics, including study design, cancer type, chemotherapy regimens, and antiemetic interventions, were summarized. Thematic synthesis focused on key clinical outcomes, including CR rates across different evaluation periods, particularly the delayed (24-120 h) and long-delayed (>120 h) phases. Where appropriate, the ranges of the reported outcomes are presented to reflect the variability across studies.

To provide an exploratory, hypothesis-generating quantitative summary of CR rates across studies, meta-analyses of proportions were carried out separately for the acute (0-24 h), delayed (24-120 h), and long-delayed (>120 h) phases. Pooled CR rates with 95% confidence intervals (CIs) were estimated using a random-effects model with the DerSimonian–Laird method to account for interstudy heterogeneity. Although alternative approaches, such as restricted maximum likelihood estimation or Hartung–Knapp adjustments, may provide more conservative estimates, the DerSimonian–Laird method was selected because of its simplicity and widespread use in meta-analyses of proportions, while allowing larger studies to contribute more substantially to the pooled estimates. No studies reported zero-event outcomes; therefore, continuity correction was not required.

Statistical heterogeneity was assessed using Cochran’s Q test and the I² statistic, with I² values of >50% considered indicative of substantial heterogeneity.

Exploratory subgroup analyses were carried out using stratified comparisons according to study-level characteristics, including study design (phase II versus phase III) and antiemetic regimen, to explore potential sources of heterogeneity. These analyses were intended for hypothesis generation rather than confirmatory inference. Univariable meta-regression was carried out to explore the association between the pooled CR rate and continuous study-level covariates, including mean/median age and the proportion of female participants, using random-effects models. Moreover, analyses involving continuous study-level variables (e.g. mean age and the proportion of female participants) were not carried out.

All statistical analyses were carried out using Stata version 18.0BE (StataCorp LLC, College Station, TX). Statistical significance for exploratory subgroup comparisons was defined as a two-sided P value of <0.05.

Quality assessment

Given the scoping review design, a formal risk-of-bias or quality assessment was not undertaken, in accordance with PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines.

Results

Study selection and characteristics

After comprehensive screening, 26 prospective studies were included in the final temporal mapping analysis (Figure 1). These included 15 phase III RCTs and 11 phase II studies (including randomized and single-arm designs). Among the included studies, five (19%) were multinational. Eleven (42%) studies originated in Japan, two (8%) in the United States, and two (8%) in India, whereas one study each (4%) was conducted in Canada, Switzerland, Thailand, Italy, China, and South Africa. The total number of patients across all the included studies was 10 858. The proportion of female patients in the overall population was 67.4%. All the studies evaluated the antiemetic efficacy of MEC. Across all the included studies, the acute phase was uniformly defined as 0-24 h after chemotherapy initiation. The delayed phase was consistently defined as 24-120 h (or 25-120 h). Notably, 21 of the 26 studies (80.8%) restricted the evaluation to 120 h (Table 1 and Figure 2).^14,16,25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48

Figure 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow diagram. This flow diagram illustrates the process of study identification, screening, eligibility assessment, and inclusion in the present scoping review of antiemetic efficacy in MEC. Records were identified through a structured PubMed/Cochrane Library/Medical Online/Ichushi-Web search and manual screening of reference lists and contemporary guideline documents. After removing duplicates, the titles and abstracts were screened for relevance. Full-text articles were assessed for eligibility on the basis of predefined inclusion criteria, including a prospective study design, an explicit definition of delayed-phase outcomes, and the reporting of CR. Twenty-six studies met the eligibility criteria for inclusion in the scoping review.

CR, complete response; MEC, moderately emetogenic chemotherapy.

Table 1.

Delayed (24-120 h) and extended (>120 h) CR rates in included MEC studies (N = 26)

Regimen category	First author (year)	Reference	Study design	MEC regimen	Antiemetic regimen	Delayed definition	Emetic events occurring within 120 h in the intervention group (%)	Delayed CR (%)	Extended definition	Extended CR (%)
5-HT3RA	Gralla et al. (2003)	25	Phase III RCT	Broad MEC	PALO versus OND	24-120 h	-	74.1 versus 55.1	Not evaluated	-
5-HT3RA	Eisenberg et al. (2003)	26	Phase III RCT	Broad MEC	PALO versus dolasetron	24-120 h	≃30% (nausea) on day 5	54–56.6 versus 38.7	Not evaluated	-
NK1RA triplet	Warr et al. (2005)	27	Phase III RCT	AC MEC	APR + OND + DEX	24-120 h	-	55 versus 49	Not evaluated	-
NK1RA triplet	Rapoport et al. (2010)	28	Phase III RCT	Broad MEC	APR + OND + DEX	25-120 h	-	AC, 70.4 versus 59.8; non-AC, 84.5 vs 73.8	Not evaluated	-
NK1RA triplet	Weinstein et al. (2016)	29	Phase III RCT	Non-AC MEC	Fosaprepitant + 5-HT3RA + DEX	25-120 h	-	83.9 versus 75.1	Not evaluated	-
NK1RA triplet	Hesketh et al. (2016)	30	Phase III RCT	Carboplatin MEC	Rolapitant + 5-HT3RA + DEX	>24-120 h	-	82.3 vs 65.6	Not evaluated	-
NK1RA triplet	Schwartzberg et al. (2015)	31	Phase III RCT	Non-AC MEC (subgroup)	Rolapitant + 5-HT3RA + DEX	>24-120 h	-	AC, 67 versus 60; non-AC, 76 versus 64	Not evaluated	-
NK1RA triplet	Nishimura et al. (2015)	14	Phase III RCT	Oxaliplatin-based MEC	APR/fosaprepitant + 5-HT3RA + DEX	24-120 h	-	85.0 versus 75.4	Not prospectively quantified	-
NK1RA triplet	Yahata et al. (2016)	32	Phase III RCT	TC (carboplatin + paclitaxel)	APR + 5-HT3RA + DEX versus placebo	24-120 h	-	63.6 versus 49.3	Not evaluated	-
NK1RA triplet	Cheng et al. (2022)	33	Phase III RCT	mFOLFOX6	APR + PALO versus DEX + PALO	25-120 h	≃10% (significant nausea) on day 5	66.7 versus 60.3	Not evaluated	-
NK1RA triplet (phase II)	Arpornwirat et al. (2009)	34	Phase II RCT	Broad MEC	Casopitant + OND + DEX	24-120 h	-	84.2 versus 69.4	Not evaluated	-
NK1RA triplet (phase II)	Kusagaya et al. (2015)	35	Phase II RCT	Carboplatin	APR + PALO + DEX	24-120 h	-	80.5 versus 76.9	Not evaluated	-
NK1RA triplet (phase II)	Maehara et al. (2015)	16	Phase II RCT	Carboplatin	APR + PALO + DEX	24-120 h	-	100 versus 66.7	Not evaluated	-
NK1RA triplet (phase II)	Tanioka et al. (2013)	36	Phase II RCT	Carboplatin MEC (selected)	APR + GRA + DEX	25-120 h	-	62.2 versus 52.1	Not evaluated	-
NK1RA triplet (phase II)	Kaushal et al. (2015)	37	Phase II RCT	Docetaxel + carboplatin + 5-FU	PALO + DEX + APR	Days 2-5	-	83.3 versus 53.3	Not evaluated	-
NK1RA triplet (phase II)	Iihara et al. (2023)	38	Phase II RCT	Trastuzumab deruxtecan	APR + GRA + DEX versus GRA + DEX	24-120 h	≃20% (nausea) on day 5	75.0 versus 36.8	24-168 h	70.0 versus 31.6
NK1RA triplet (phase II)	Aoyama et al. (2025)	39	Phase II RCT	Trastuzumab deruxtecan	PALO + DEX + APR versus PALO + DEX	Days 2-21	-	41.4 versus 37.9	-	-
NEPA	Aapro et al. (2014)	40	Phase III RCT	AC MEC	NEPA versus PALO	25-120 h	-	76.9 versus 69.5	Not evaluated	-
NEPA	Zelek et al. (2023)	41	Phase III RCT	Non-AC MEC	NEPA versus APR	24-144 h	-	89.9 versus 83.3	0-144 h	77.1 vs 57.8
Olanzapine	Ostwal et al. (2024)	42	Phase III RCT	OXA/carboplatin/IRI MEC	NK1 + PALO + DEX ± OLZ	25-120 h	-	92 versus 83	Not evaluated	-
Olanzapine	Iihara et al. (2020)	43	Phase II single-arm	Carboplatin AUC ≥4	OLZ + NK1 + 5-HT3RA + DEX	24-120 h	≃20% (nausea) on day 5	80.7	24-168 h	75.7
Olanzapine	Sakai et al. (2021)	44	Phase II single-arm	Carboplatin AUC ≥5	Low-dose olanzapine + GRA + DEX	24-120 h	6% (nausea) on day 5	100	94	-
Olanzapine	Sakai et al. (2025)	45	Phase II RCT	Trastuzumab deruxtecan	OLZ + 5-HT3RA + DEX versus 5-HT3RA + DEX	24-120 h	≃30% (nausea) on day 5	70.0 versus 56.1	120-504 h	63.9
Mirtazapine	Iihara et al. (2024)	46	Phase II single-arm	Carboplatin AUC ≥4	Mirtazapine + 5-HT3RA + DEX	24-120 h	≃10% (nausea) on day 5	83.3	Not evaluated	-
Steroid-sparing	Aapro et al. (2010)	47	Phase III RCT	Breast MEC	PALO + DEX day 1 versus days 1-3	24-120 h	-	62.3 versus 65.8	Not evaluated	-
Steroid-sparing	Celio et al. (2011)	48	Phase III RCT	Broad MEC	PALO + DEX day 1 versus days 1-3	24-120 h	-	68.7 versus 77.7	Not evaluated	-

AC, anthracycline–cyclophosphamide; APR, aprepitant; AUC, area under the curve; CR, complete response; DEX, dexamethasone; 5-FU, 5-fluorouracil; GRA, granisetron; 5-HT₃RA, 5-hydroxytryptamine-3 receptor antagonist; IRI, irinotecan; MEC, moderately emetogenic chemotherapy; mFOLFOX6, modified FOLFOX6 (fluorouracil, leucovorin, and oxaliplatin); NEPA, netupitant–palonosetron fixed combination; NK1, neurokinin-1; NK1RA, neurokinin-1 receptor antagonist; OLZ, olanzapine; OND, ondansetron; OXA, oxaliplatin; PALO, palonosetron; RCT, randomized controlled trial; TC, total control.

Figure 2.

CR rates across antiemetic strategies and time windows in MEC trials. This figure summarizes the CR rates reported in prospective clinical studies of MEC stratified according to antiemetic strategy and assessment time window. The horizontal axis represents the time after chemotherapy administration and is divided into acute (0-24 h), delayed (>24-120 h), and long-delayed (>120 h) phases. Horizontal bars represent the treatment arm used in the clinical study. CR was defined as no emesis and no use of rescue medication, unless otherwise specified in the original study. Numerical values within the bars indicate the reported CR rates for each phase. Antiemetic strategies are grouped according to category on the vertical axis: (i) 5-HT₃RA + DEX, (ii) NK1RA-containing triplet regimens, (iii) NEPA-containing triplet regimens, (iv) olanzapine-containing regimens, (v) mirtazapine-containing regimens, and (vi) steroid-sparing strategies. Symbols indicate the study phase: green diamonds represent phase III trials and blue diamonds represent phase II studies. Most trials report CR through 120 h only. Long-delayed phase data (>120 h) are shown only where explicitly reported (e.g. 0-144 h or 24-168 h). The figure highlights the predominance of the 120-h cut-off across antiemetic strategies despite advances in pharmacological regimens.

AC, anthracycline–cyclophosphamide; APR, aprepitant; AUC, area under the curve; Carbo, carboplatin; CR, complete response; DEX, dexamethasone; FosAPR, fosaprepitant; 5-FU, 5-fluorouracil; GRA, granisetron; 5-HT₃RA, 5-hydroxytryptamine-3 receptor antagonist; IRI, irinotecan; L-OHP, oxaliplatin and oxaliplatin; mFOLFOX6, modified FOLFOX6 (fluorouracil, leucovorin); MEC, moderately emetogenic chemotherapy; NEPA, netupitant–palonosetron fixed combination; NK1, neurokinin-1; NK1RA, neurokinin-1 receptor antagonist; OLZ, olanzapine; OND, ondansetron; PALO, palonosetron; TC, total control.

Acute phase (0–24 h)

The pooled CR rate in the acute phase was 87% (95% CI 85% to 90%; I² = 93.73%; Supplementary Figure S1, available at https://doi.org/10.1016/j.esmoop.2026.108582). Studies reporting 5-HT₃RAs with DEX reported a CR of 82% (95% CI 77% to 86%). In studies in which NK1RAs were added, the CR rate was 91% (95% CI 88% to 93%). When olanzapine was added, the CR rate was 96% (95% CI 94% to 98%). Exploratory subgroup analyses suggested differences between phase II and III studies (P < 0.01; Supplementary Figure S2, available at https://doi.org/10.1016/j.esmoop.2026.108582). Phase II studies had higher CR rates (93%; 95% CI 90% to 96%; I² = 70.69%; Supplementary Figure S2, available at https://doi.org/10.1016/j.esmoop.2026.108582) compared with phase III studies (84%; 95% CI 80% to 87%; I² = 93.73%; Supplementary Figure S2, available at https://doi.org/10.1016/j.esmoop.2026.108582). Exploratory subgroup analyses suggested higher pooled CR rates with increasing age (P < 0.01) and in studies including fewer female patients (P < 0.01) (Supplementary Figures S3 and S4, available at https://doi.org/10.1016/j.esmoop.2026.108582).

Delayed phase (24-120 h)

Several NK1RAs, including aprepitant-, rolapitant-, and NTP-based combinations, were evaluated in patients receiving MEC. Phase II randomized studies of carboplatin- or oxaliplatin-based MEC showed similarly variable delayed CR rates. However, even in contemporary trials incorporating olanzapine or optimized NK1RA-based regimens, the evaluation was confined to 120 h.

The pooled CR rate in the delayed phase was 71% (95% CI 68% to 75%; I² = 93.22%; Supplementary Figure S5, available at https://doi.org/10.1016/j.esmoop.2026.108582). Exploratory subgroup analyses suggested differences in pooled CR rates among antiemetic regimens (P < 0.01; Supplementary Figure S5, available at https://doi.org/10.1016/j.esmoop.2026.108582). Moreover, exploratory subgroup analyses also suggested higher pooled CR rates with increasing age (P = 0.02; Supplementary Figure S6, available at https://doi.org/10.1016/j.esmoop.2026.108582) and lower pooled CR rates in studies including higher proportions of female patients (P < 0.01; Supplementary Figure S7, available at https://doi.org/10.1016/j.esmoop.2026.108582). Early trials of 5-HT₃RAs with DEX reported delayed CR rates of 63% (95% CI 58% to 67%; I² = 89.74%; Supplementary Figure S5, available at https://doi.org/10.1016/j.esmoop.2026.108582). With the introduction of NK1RA-containing triplet regimens, delayed CR rates generally reached 79% (95% CI 74% to 83%; I² = 90.70%; Supplementary Figure S5, available at https://doi.org/10.1016/j.esmoop.2026.108582). Steroid-sparing regimens using 5-HT₃RAs yielded a delayed CR rate of 66% (95% CI 59% to 72%; I² = 30.29%; Supplementary Figure S5, available at https://doi.org/10.1016/j.esmoop.2026.108582). Olanzapine-containing regimens with 5-HT₃RAs and DEX reached CR rates of 83% (95% CI 55% to 99%; I² = 92.3%; Supplementary Figure S5, available at https://doi.org/10.1016/j.esmoop.2026.108582). Olanzapine combined with NK1RA-containing quadruplet therapy demonstrated a pooled delayed CR rate of 88% (95% CI 75% to 96%; I² = 81.70%; Supplementary Figure S5, available at https://doi.org/10.1016/j.esmoop.2026.108582). Phase II studies had higher CR rates (75%; 95% CI 70% to 80%; I² = 77.60%; Supplementary Figure S8, available at https://doi.org/10.1016/j.esmoop.2026.108582) compared with phase III studies (69%; 95% CI 65% to 74%; I² = 95.36%; Supplementary Figure S8, available at https://doi.org/10.1016/j.esmoop.2026.108582). Exploratory subgroup analyses did not identify differences between phase II and phase III studies (P = 0.12; Supplementary Figure S8, available at https://doi.org/10.1016/j.esmoop.2026.108582).

Long-delayed phase (>120 h)

Five prospective studies extended CINV assessment beyond the conventional 120-h timeframe. All AC-based studies were limited to the conventional 24-120 h delayed-phase assessment and did not evaluate outcomes beyond 120 h. The extended-assessment studies comprised one phase III non-AC MEC trial assessing outcomes through 144 h,³⁶ one carboplatin phase II study, and one trastuzumab deruxtecan phase II study assessing outcomes through 168 h.^33,40 Aoyama et al.³⁴ assessed trastuzumab deruxtecan–associated CINV over days 2–21 and Sakai et al.⁴⁴ specifically evaluated outcomes during the 120-504 h period.

In this study, the extended overall CR (0-144 h) differed from the CR reported during the conventional 0-120 h window, indicating that evaluation beyond 120 h may influence treatment comparisons. Additionally, a single-arm phase II study evaluated CR for 168 h in carboplatin regimens with an area under the curve of ≥4 ,⁴⁰ reporting a modest decline in CR when extending from 120 to 168 h. No traditional phase III RCT conducted before 2023 prospectively quantified CR beyond 120 h. For trastuzumab deruxtecan, studies evaluating CINV over extended observation periods of ≤168 or ≤504 h after chemotherapy were identified.^33,34,44

Overall, the CR rate across studies evaluating the long-delayed phase was 62% (95% CI 46% to 76%; I² = 91.70%; Supplementary Figure S9, available at https://doi.org/10.1016/j.esmoop.2026.108582). Exploratory subgroup analyses suggested differences among antiemetic regimens (P < 0.01; Supplementary Figure S9, available at https://doi.org/10.1016/j.esmoop.2026.108582). Studies evaluating 5-HT₃RAs with DEX have reported a CR rate of 41% (95% CI 32% to 49%; I² = 0.000%; Supplementary Figure S9, available at https://doi.org/10.1016/j.esmoop.2026.108582). Studies that included NK1RAs reported a CR rate of 74% (95% CI 55% to 90%; I² = 89.55%; Supplementary Figure S9, available at https://doi.org/10.1016/j.esmoop.2026.108582). Exploratory subgroup analyses suggested differences according to study design (P < 0.01; Supplementary Figure S10, available at https://doi.org/10.1016/j.esmoop.2026.108582). Studies with a phase III design had a higher CR rate (87%; 95% CI 80% to 93%; I² = 48.38%; Supplementary Figure S10, available at https://doi.org/10.1016/j.esmoop.2026.108582) compared with those with a phase II design (53%; 95% CI 40% to 66%; I² = 79.00%; Supplementary Figure S10, available at https://doi.org/10.1016/j.esmoop.2026.108582). No differences according to study-level age (P = 0.74; Supplementary Figure S11, available at https://doi.org/10.1016/j.esmoop.2026.108582) or sex (P = 0.97; Supplementary Figure S12, available at https://doi.org/10.1016/j.esmoop.2026.108582) were identified in exploratory subgroup analyses.

The delayed phase was defined according to each original study and typically ranged from 24 h to 120 h or 25 h to 120 h after the initiation of chemotherapy. The long-delayed phase refers to an evaluation beyond 120 h (e.g. 0-144 h or 24-168 h). CR was generally defined as no emesis or the use of rescue medication; however, some trials required the absence of significant nausea (see individual trial definitions). Most randomized phase III trials limited the evaluation to 120 h; only selected studies extended the evaluation beyond this timeframe. Where two CR percentages are presented, the values correspond to the experimental and control arms, respectively. Single-arm phase II studies were included to contextualize long-delayed phase reporting; however, they should not be directly compared with randomized trials. Chemotherapy regimens are presented according to the emetogenic risk classification used in the original studies. Because emetogenic classifications have evolved over time and may differ among contemporary guidelines, some regimens or agents may now be classified or managed differently. In particular, AC regimens are now generally classified as highly emetogenic, whereas the classification and recommended antiemetic management of carboplatin-based chemotherapy and trastuzumab deruxtecan are not entirely uniform across contemporary guidelines. AC-based studies did not evaluate outcomes beyond 120 h; hence, they did not contribute to the long-delayed CINV evidence base.

Discussion

The primary objective of this scoping review was not to compare the efficacy of antiemetic regimens but rather to examine how delayed and long-delayed CINV have been defined and assessed in prospective clinical trials involving MEC and to identify gaps in the current evidence.

In this scoping review, we systematically mapped the definitions and reported delayed and long-delayed phases in prospective clinical trials of MEC. Despite substantial advances in antiemetic pharmacotherapy, including long-acting 5-HT₃RAs, NK1RAs, fixed NK1RA–5-HT₃RA combinations, and olanzapine-containing regimens, the temporal framework for assessing delayed CINV has remained largely fixed at 120 h. Across the 26 included studies, delayed CINV was almost uniformly defined as occurring within 24-120 h after chemotherapy administration. This predominance of the 120-h evaluation window suggests that the true burden of CINV beyond 120 h may be systematically underestimated. Taken together, these findings highlight that long-delayed CINV is common and should be more clearly recognized, systematically evaluated, and addressed in future studies and clinical practice. However, evidence beyond 120 h remains limited because only a few prospective studies have systematically evaluated this time period. Nevertheless, findings from an individual patient data meta-analysis by Chow et al.²² suggest that a proportion of patients may experience long-delayed CINV, supporting the clinical relevance of this time period.

The principal finding of this scoping review is that prospective phase II/III antiemetic trials for MEC have almost uniformly limited CINV assessment to the conventional 24-120-h delayed phase, leaving long-delayed CINV (>120 h) largely uncharacterized. Although only a limited number of studies extended outcome assessment beyond 120 h, these studies consistently suggested a further decline in CR rates, with a sample size–weighted overall CR rate of 62%. Although this observation should be interpreted cautiously given the limited evidence base, it raises the possibility that the conventional 120-h assessment window may underestimate the burden of persistent CINV. Consistent with current guideline-based strategies for HEC, regimens incorporating NK1RA in addition to 5-HT₃RAs and DEX were associated with numerically higher CR rates in both the acute and delayed phases, whereas regimens including olanzapine showed a similar trend in the acute phase.

These findings reinforce the importance of multidrug prophylaxis targeting distinct neurotransmitter pathways. The mechanisms underlying long-delayed CINV are likely to be more complex than those responsible for acute or conventional delayed CINV. Although serotonin and substance P remain key mediators, other neurotransmitter systems, including dopamine, histamine, and muscarinic pathways, may contribute to persistent nausea beyond 120 h after chemotherapy. This broader neuropharmacological involvement may partly explain the favorable efficacy of olanzapine, which antagonizes multiple neurotransmitter receptors, including dopamine, serotonin, histamine, and muscarinic receptors. Further mechanistic and clinical studies are warranted to clarify the pathophysiology of long-delayed CINV and identify optimal preventive strategies. In addition to receptor-mediated mechanisms involving dopamine, histamine, and muscarinic pathways, regimen-specific pharmacokinetic and pharmacodynamic factors may also contribute to long-delayed nausea and vomiting. This consideration may be particularly relevant for antibody–drug conjugates (ADCs) such as trastuzumab deruxtecan (T-DXd). T-DXd delivers the membrane-permeable topoisomerase I inhibitor payload DXd through a cleavable linker and the apparent elimination half-lives of fam-trastuzumab deruxtecan-nxki and released DXd after administration are ∼5.4–5.7 days and ∼5.4–6.1 days, respectively. Therefore, nausea and vomiting occurring beyond 120 h after trastuzumab deruxtecan may not only be fully explained by the conventional time course established for cisplatin-induced delayed CINV but may also reflect sustained ADC/payload exposure and prolonged emetogenic signaling. This hypothesis requires further investigation through prospective studies incorporating time-resolved symptom assessment and pharmacokinetic/pharmacodynamic analyses.

Importantly, our analysis extends the existing evidence by demonstrating that CR rates continue to decline when the observation period is extended beyond 120 h. Limited but consistent data from studies evaluating outcomes up to 144–168 h indicate that conventional endpoints may overestimate antiemetic efficacy. This is particularly relevant in the context of newer agents, such as ADCs, for which prolonged emetogenicity has been reported. These findings support the need to reconsider standard assessment windows and incorporate longer follow-up durations into future clinical trials and guideline recommendations. The inclusion of ADC studies reflects the evolving landscape of emetogenic chemotherapy and the increasing clinical relevance of prolonged observation periods rather than an assumption that their emetogenic profiles are equivalent to those of conventional MEC regimens.

Exploratory subgroup analyses suggested that studies including older patients tended to report higher pooled CR rates, whereas those with higher proportions of female patients tended to report lower pooled CR rates in the acute and delayed phases, in line with established risk factors for CINV.49, 50, 51, 52, 53 However, these associations were not observed in the long-delayed phase, potentially reflecting the limited sample size or differing mechanisms underlying very late–phase symptoms.

Only five prospective studies extended evaluation beyond the conventional 120-h timeframe: one phase III trial evaluated CINV through 144 h,³⁶ three phase II studies extended the assessment beyond 120 h,^33,34,40 and one with the longest observation period reached 504 h.⁴⁴ Some ADCs, including trastuzumab deruxtecan, contain payloads that can induce CINV over prolonged periods. Consequently, the evaluation of CINV incidence beyond the conventional definition of delayed CINV may be increasingly important in patients receiving these agents.⁵⁴ Recent evidence has highlighted the need for extended observation periods to better characterize CINV associated with ADC therapies.²¹ In HEC, trials evaluating long-acting NK1RAs, such as fosnetupitant, have prospectively assessed outcomes for 168 h.⁵⁵ These data demonstrate that extended evaluation beyond 120 h is feasible within randomized clinical trial frameworks and may provide clinically meaningful insights into late-phase symptom control. The limited data available beyond 120 h in MEC indicate that clinically relevant symptoms may continue into the long-delayed phase. Supporting this concern, a recent individual patient data meta-analysis specifically evaluated long-delayed CINV and reported that 24% of patients receiving MEC experienced long-delayed nausea and 6% experienced long-delayed vomiting.²² These findings are consistent with the results of the present scoping review, in which the incidence rate of nausea during the long-delayed phase ranged from ∼20% to ∼30%, further emphasizing the importance of evaluating long-delayed CINV. Notably, the severity of nausea during the long-delayed phase was comparable with that observed during the conventional delayed phase. These findings suggest that the clinical burden after 120 h is not negligible and may warrant systematic evaluation in future prospective trials. In the pragmatic NEPA trial extending to 144 h,³⁶ the extended overall CR differed from the CR reported during the conventional 0-120 h window, indicating that treatment effects may not be fully captured within the traditional window. Similarly, a phase II study⁴⁰ with a 168-h window demonstrated a modest decline in CR when the assessment was extended beyond 120 h. Furthermore, across studies, the proportion of patients experiencing emetic events at 120 h among patients administered a NK1RA-containing triplet regimen was ∼20%, suggesting that emetic events may persist beyond 120 h in a subset of patients.

Given that platinum-based regimens, such as oxaliplatin and carboplatin, are commonly administered in outpatient settings, late symptoms may remain under-recognized.

Extending symptom monitoring beyond 120 h may also help improve the understanding of late-phase CINV and inform supportive care strategies in outpatient chemotherapy settings.

Implications for MASCC/ESMO and ASCO updates

The findings of this review have potential implications for future updates to international antiemetic guidelines, including the MASCC/ESMO and ASCO recommendations. Current guidelines primarily define the delayed phase of CINV as occurring within 120 h of chemotherapy and the corresponding clinical trial endpoints are aligned with this timeframe.

Given the emerging evidence suggesting that CINV may extend beyond 120 h in a subset of patients, extending the evaluation period in both clinical trials and practice should be considered. Incorporating longer observation windows may improve the accuracy of antiemetic efficacy assessment and better reflect patient experiences, particularly in outpatient settings.

Additionally, guideline panels should consider whether specific antiemetic strategies with prolonged pharmacological activity, such as long-acting NK1RAs, can play a role in addressing long-delayed symptoms. However, given the limited prospective evidence currently available, further studies are required to establish formal recommendations.

Another potential factor contributing to the persistent use of the 120-h boundary is the design of commonly used patient-reported outcome tools for CINV. Many clinical trials rely on patient diaries or validated instruments, such as the MASCC Antiemesis Tool, which were originally developed to capture symptoms during the first 5 days following chemotherapy. Consequently, the traditional 120-h evaluation window may partly reflect the structure of these assessment tools rather than a biologically defined boundary of symptom occurrence. Future studies may need to reconsider both trial design and symptom assessment tools to capture long-delayed CINV more comprehensively.⁵⁶

When considering the pharmacological mechanisms based on central neurokinin-1 (NK1) receptor occupancy, data on the time-dependent brain NK1 receptor occupancy of aprepitant and NTP should be considered. Detailed time-dependent brain NK1 receptor occupancy data are available for aprepitant- and NTP-based agents, whereas comparable positron emission tomography studies on other NK1RAs are limited. When aprepitant is administered at a dose of 150 mg intravenously or 165 mg orally, the brain NK1 receptor occupancy at 120 h after administration is ∼50%.⁵⁷ In contrast, 235 mg of intravenous fosnetupitant maintains higher receptor occupancy, with a brain NK1 receptor occupancy of 68.2% at 120 h and 60.3% at 168 h after administration.⁵⁸ In the present review, the incidence rate of emetic events remained ∼20%-30% within 120 h after chemotherapy initiation even with NK1RA-based triplet antiemetic regimens and a proportion of patients newly developed emetic events around the 120-h time point. These findings suggest that new-onset CINV may occur beyond 120 h given that NK1 receptor occupancy declines over time. Notably, NTP maintains a relatively prolonged central NK1 receptor occupancy compared with aprepitant. This pharmacological characteristic may contribute to improved CINV control during the long-delayed phase. NTP, a novel NK1RA, has demonstrated noninferiority to aprepitant in combination with PALO and DEX for HEC with a favorable safety profile.^55,59 Its long elimination half-life of ∼144 h⁵⁵ and sustained NK1 receptor occupancy⁵⁹ suggest potential prolonged antiemetic activity in the delayed phase.

In Japan, a prospective observational study is currently underway to evaluate the antiemetic efficacy of NK1RA (fosnetupitant)-containing triplet regimens for platinum-based chemotherapy through 168 h.⁶⁰ Such real-world data will be critical for clarifying the incidence and clinical significance of long-delayed CINV in MEC. Accumulating evidence from observational and pragmatic studies may ultimately indicate whether routine evaluation beyond 120 h should be incorporated into future clinical trial designs and guideline recommendations.

In conclusion, although the pharmacological strategies for MEC have substantially evolved, the temporal definition of delayed CINV remains largely unchanged. Evidence from HEC trials demonstrates that an extended evaluation period of 168 h is feasible. Extending the systematic assessment of CINV beyond 120 h in MEC may represent the next necessary step in optimizing antiemetic care.

This study has some limitations. Firstly, the literature search was limited to PubMed, Cochrane Library, Medical Online, Ichushi-Web, and manual screening of guideline-related references. Secondly, only five studies evaluated the extended delayed phase (>120 h). Thirdly, the studies included in the pooled analyses differed in CR definitions, observation periods, chemotherapy regimens, and patient populations; hence, they should be interpreted with caution. The quantitative synthesis of these studies was intended to provide a descriptive overview rather than a definitive estimate of antiemetic efficacy. Fourthly, substantial heterogeneity was observed across analyses, which may have limited the interpretability of the pooled estimates. Therefore, these findings should be considered exploratory and interpreted with caution. Moreover, prediction intervals were not calculated, which may have limited our ability to assess the expected range of effects in future studies. Formal assessment of publication bias (e.g. funnel plots or Egger’s test) was not carried out because of the limited number of studies and substantial heterogeneity, which may reduce the reliability of such analyses. Lastly, because emetogenic risk classifications have evolved over time and differ among current international guidelines, some chemotherapy regimens included in this review may be categorized differently in contemporary practice. However, studies were selected according to the classification applied at the time of trial design to preserve the historical context of prospective evidence.

Conclusion

Delayed CINV in MEC has been uniformly assessed within 120 h in most prospective trials. Despite the feasibility of HEC studies, evidence beyond this timeframe is limited. Extending the evaluation beyond 120 h may be necessary to fully characterize long-delayed CINV and optimize antiemetic strategies.