Nausea and emesis remain significant problems of chemotherapy despite prophylaxis with 5-hydroxytryptamine-3 antiemetics

A University of Rochester James P. Wilmot Cancer Center Community Clinical Oncology Program study of 360 cancer patients treated in the community


  • Jane T. Hickok M.D., M.P.H.,

    Corresponding author
    1. University of Rochester Cancer Center (URCC) Community Clinical Oncology Program (CCOP) Research Base, University of Rochester James P. Wilmot Cancer Center, Rochester, New York
    • University of Rochester James P. Wilmot Cancer Center, 601 Elmwood Avenue, P.O. Box 704, Rochester, NY 14642
    Search for more papers by this author
    • Fax: (585) 461-5601

  • Joseph A. Roscoe Ph.D.,

    1. University of Rochester Cancer Center (URCC) Community Clinical Oncology Program (CCOP) Research Base, University of Rochester James P. Wilmot Cancer Center, Rochester, New York
    Search for more papers by this author
  • Gary R. Morrow Ph.D., M.S.,

    1. University of Rochester Cancer Center (URCC) Community Clinical Oncology Program (CCOP) Research Base, University of Rochester James P. Wilmot Cancer Center, Rochester, New York
    Search for more papers by this author
  • David K. King M.D.,

    1. Greater Phoenix Community Clinical Oncology Program (CCOP), Phoenix, Arizona
    Search for more papers by this author
  • James N. Atkins M.D.,

    1. Southeast Cancer Control Consortium, Winston-Salem, North Carolina
    Search for more papers by this author
  • Tom R. Fitch M.D.

    1. Mayo Clinic Scottsdale Community Clinical Oncology Program (CCOP), Scottsdale, Arizona
    Search for more papers by this author



Clinical reports suggest that nausea remains a side effect of chemotherapy despite widespread use of serotonin receptor antagonists. This study summarized the frequency, timing, and intensity of postchemotherapy nausea for patients receiving doxorubicin, cisplatin, or carboplatin.


Three hundred sixty chemotherapy-naïve patients (73% female) were enrolled in a study testing the ability of an information intervention to reduce nausea. Of these, 322 subjects completed the Morrow Assessment of Nausea and Emesis (MANE), as well as a 5-day self-report diary, at Cycle 1 (300 subjects completed the MANE and self-report diary at Cycle 2). All patients received a 5-hydroxytryptamine-3 receptor antagonist (ondansetron) with dexamethasone on the day of treatment.


Seventy-six percent of the patients developed nausea during the 5-day period, beginning with the Cycle 1 infusion, and 73% of patients reported delayed nausea (DN) during Days 2–5. The proportions were similar during Cycle 2. Fifty-five percent of patients described their DN as being of moderate or greater intensity compared with 28% of patients who described acute nausea. Carboplatin was less likely to cause DN than either of the other agents (56% of 106 patients compared with 75% of 47 receiving cisplatin and 83% of 169 taking doxorubicin). The mean peak DN severity was 4.34 (range, 1–7) for doxorubicin, which was significantly higher than the mean peak value for carboplatin (3.66) but was not significantly different from the mean peak value for cisplatin (4.26). Eighteen percent of patients did not experience nausea until Day 3 or later.


Despite prophylaxis with ondansetron, the majority of patients receiving one of these common chemotherapy agents experienced nausea. The frequency of DN was nearly twice that of acute nausea. Results show the need for continued development of antiemetics that are effective against DN. Cancer 2003;97:2880–6. © 2003 American Cancer Society.

DOI 10.1002/cncr.11408

For many patients who undergo chemotherapy for cancer, nausea is the most disturbing side effect of their treatment experience.1, 2 Twenty years ago, emesis was the number one scourge of chemotherapy.3 This is no longer the case because of the widespread integration into clinical oncology practice of the 5-hydroxytryptamine-3 (5-HT3) receptor antagonist class of antiemetics. Antiemetics include such drugs as ondansetron (Zofran, GlaxoSmithKline, Research Triangle Park, NC) and granisetron (Kytril, Roche Laboratories, Nutley, NJ).

Chemotherapy-related nausea is characterized as either postchemotherapy nausea, which occurs after chemotherapy infusion, or anticipatory nausea, which is experienced by approximately 25% of patients in anticipation of receiving chemotherapy. Postchemotherapy nausea can be grouped into acute nausea, which develops within 18–24 hours after infusion, and delayed nausea (DN), which occurs on Day 2 or later after treatment. Different mechanisms of action may account for the various types of chemotherapy-related nausea.4–6

Of particular concern is the occurrence of DN. Although serotonin receptor antagonist drugs are more effective against nausea than previously available antiemetic drugs, evidence suggests that they are less effective in controlling DN than acute nausea.7 Delayed nausea is described most often in conjunction with the administration of highly emetogenic chemotherapy drugs such as doxorubicin and cisplatin. After the administration of cisplatin, 50–80% of patients experienced DN, the frequency being related to the dose used.8

This study presents detailed information on the frequency, severity, and persistence of acute nausea, DN, and emesis over the first two cycles of chemotherapy. We evaluated 322 patients who received a serotonin receptor antagonist antiemetic on the day of treatment with doxorubicin, cisplatin, or carboplatin.



Study subjects were clinic outpatients at 18 private medical oncology practice groups. These groups were grantees of the National Cancer Institute's Community Clinical Oncology Program (CCOP) and were members of the University of Rochester James P. Wilmot Cancer Center (URCC) CCOP research base between January 12, 1998 and September 25, 2000. Adult cancer patients who were 18 years of age or older were eligible for participation if they were scheduled to receive outpatient chemotherapy for the first time that comprised a regimen containing cisplatin (Baxter Healthcare, New Providence, NJ), carboplatin (Bristol Myers Squibb, Princetion, NJ), or doxorubicin (Baxter Healthcare).

Design and Procedures

Data were compiled from a two-arm randomized clinical trial comparing the effect of a specific education intervention (designed to provide an enhanced positive expectation for the efficacy of antiemetic medication) plus standard educational materials (pamphlets entitled “Chemotherapy and You” [NIH Pub. No. 99-1136, 1999] and “Helping Yourself During Chemotherapy” [NIH Pub. No. 94-3701, 1994]) with the standard educational pamphlets alone on the frequency and severity of postchemotherapy nausea and emesis. Randomization to the two study arms was stratified by CCOP site and by chemotherapy agent (cisplatin ≥ 60 mg/m2 vs. cisplatin < 60 mg/m2 vs. carboplatin vs. doxorubicin). The protocol was approved by the institutional review board of the URCC and of each participating site before patient enrollment. Each patient signed a consent form approved by the appropriate review board before being enrolled on the study.

Ondansetron (24 mg orally or 20 mg intravenously [i.v.]) and dexamethasone (12 mg orally or 10 mg i.v.) were administered to all study participants 30 minutes before chemotherapy infusion on Day 1. Beginning on Day 2, each patient received the antiemetic regimen that comprised standard care at each participating CCOP practice site at the discretion of the treating oncologist. All other ancillary treatments needed for control of symptoms caused by the cancer or its treatment were given as needed.

Assessment of Nausea and Emesis

Patients completed a self-report diary (a 5-day home record) in which the timing and intensity of nausea and the timing and number of episodes of emesis were recorded. The diary was a modified version of the one developed by Burish et al.,9 which has been used in several multicenter studies.Each day in the diary is divided into four time periods (morning, afternoon, evening, night). Subjects reported the severity of nausea and the number of emesis episodes for each time period on the day of treatment and on the 4 following days (20 total reporting times). Severity of nausea is assessed on a 7-point semantic rating scale anchored at one end by 1 (not at all nauseated) and at the other end by 7 (extremely nauseated). A score of 1 indicates no nausea, a score of 2 signifies mild nausea, a score of 3, 4, or 5 is coded as moderate nausea, and a score or 6 or 7 denotes severe nausea. Mean nausea severity scores were calculated from the last three reporting periods on Day 1 (for acute nausea) and from the 16 reporting periods from Days 2–5 (for DN). Questionnaires were completed by patients at home after their first two chemotherapy cycles and returned by mail in a postage-paid envelope.


As there were no significant differences in the frequency or severity of acute nausea or DN by assignment to intervention or control conditions in this study, all patients who provided evaluable data were combined for analyses. Furthermore, because there were no significant differences in the severity of acute nausea or DN by cisplatin dose (29 patients received < 60 mg/m2 and 18 received ≥ 60 mg/m2), all patients treated with cisplatin were combined into one group for the purpose of data analysis.

Descriptive statistics, chi-square tests, and Student t tests, as appropriate, were used to characterize the frequency, severity, and persistence of postchemotherapy nausea and emesis. Hierarchical linear regression was used to investigate the relative contribution of gender, age, and chemotherapy agent to nausea outcomes.


Of the 360 enrolled patients, 322 (89%) provided evaluable data for Cycle 1 and 300 (83%) provided data for Cycle 2. A higher percentage of males (17 of 104 [16%]) than females (20 of 255 [8%]) failed to provide evaluable outcome data at Cycle 1 (χ2Pearson = 5.78, P = 0.02, two-sided; gender information was missing for one case). There were no gender differences in the proportion of patients providing complete data at Cycle 2. Of the 169 patients treated with doxorubicin, 142 (84%) also received cyclophosphamide, 21 (12%) received vincristine, and 12 (8%) received a regimen containing taxotere. Paclitaxel was included in the carboplatin-containing regimens that 88 of 106 patients (83%) received whereas 9 (8%) received etoposide. Fifteen (32%) of the patients randomized to the cisplatin arm also received etoposide, 12 (25%) received 5-fluorouracil, and 5 (11%) were treated with paclitaxel. A few patients in each arm received regimens containing other chemotherapy agents. Data presented below are from the first two chemotherapy cycles. Table 1 summarizes demographic and clinical information for the 322 patients who provided evaluable data at Cycle 1.

Table 1. Demographic and Clinical Characteristics for 322 Patients
  1. SD: standard deviation.

Age (yrs) 
 Mean (SD)56.6 (12.7)
 Male 87 (27%)
 Female235 (73%)
 White281 (87%)
 Black 21 (7%)
 Hispanic 10 (3%)
 Some college 86 (58%)
 High school 97 (30%)
Marital status 
 Married222 (69%)
 Not married100 (31%)
 Breast142 (44%)
 Lung 71 (22%)
 Gynecologic 29 (9%)
 Hematologic 26 (8%)
 Genitourinary 19 (6%)
 Gastrointestinal 13 (4%)
 Doxorubicin169 (52%)
 Carboplatin106 (33%)
 Cisplatin 47 (15%)

Postchemotherapy Nausea

Overall, 76% of the 322 patients experienced nausea during the period from receipt of chemotherapy through Day 5 of Cycle 1, despite antiemetic prophylaxis with a 5HT3 agent. The frequency of nausea was similar after Cycle 2, with 70% of the 300 patients reporting nausea at some point during the 5-day period. One-third of the patients reported nausea on at least 4 consecutive days of their first chemotherapy cycle and 22% reported feeling nauseated on all 5days. The data for Cycle 2 were comparable at 30% and 18%. Nausea differed by gender at Cycle 1 but not at Cycle 2, with 68% of males and 79% of females reporting nausea at some point during the 5-day period after receipt of their first cycle of chemotherapy (χ2 = 4.48, P = 0.04).

Delayed Nausea

Nearly three-fourths of the 322 patients (73%) felt nauseated at some time during the delayed period at Cycle 1 and the percentage of patients was similar (68%) at Cycle 2. Thirty-six percent of patients experienced nausea only during the delayed period of both cycles. The peak severity of nausea that occurred during the delayed period was higher than the peak severity of nausea that occurred on the day of chemotherapy. Twenty-eight percent of patients rated their worst nausea on the day of chemotherapy as moderate or severe in intensity whereas 55% of patients had nausea of at least moderate intensity at some point during Days 2–5.

Postchemotherapy Emesis

The frequency of emesis was lower than that of nausea and also decreased from the first to the second cycle. Thirty percent of patients reported emesis after Cycle 1 and 25% of patients reported emesis after Cycle 2. At Cycle 1, 36 of the 322 patients (11%) vomited on the day of treatment, 60 (19%) reported emesis only during the delayed period (Days 2–5), and 21 patients (6%) experienced emesis during both the acute and delayed periods. Comparable frequencies of emesis decreased somewhat following the second cycle. At both Cycles 1 and 2, approximately 50% of the patients who experienced emesis as a result of chemotherapy vomited more than twice. Figure 1 shows the overall frequency of acute nausea and DN and emesis by chemotherapy cycle.

Figure 1.

Profile showing progress of patients through the trial.

Acute Nausea and DN by Chemotherapy Agent

The overall frequency of nausea differed by chemotherapy agent. At Cycle 1, nausea occurred in 61% of 106 patients receiving carboplatin, in 75% of 47 patients treated with cisplatin, and in 86% of 169 patients who received doxorubicin. The data for Cycle 2 were comparable at 55%, 75%, and 78%, respectively. The percentage of patients receiving doxorubicin who experienced nausea on the day of treatment (55%) was not significantly different from the percentage of patients receiving cisplatin (42.6%). However, the percentage of patients receiving carboplatin who experienced day-of-treatment nausea (24.5%) was significantly smaller than the percentages of either of these groups (Ps < 0.05). The nausea caused by doxorubicin was more likely to be of moderate or severe intensity. For patients who reported acute nausea, mean peak nausea severity was 4.3 for patients receiving doxorubicin, which is significantly higher than the mean peak nausea severity reported by patients receiving cisplatin (3.3) or carboplatin (2.8; Ps < 0.05). Worst acute nausea severity was not significantly different for the two platinum-based chemotherapy agents.

There also were significant differences in the frequency and severity of DN by chemotherapy agent. Carboplatin was less likely to cause DN than either of the other two agents (Ps < 0.05). Fifty-six percent of the 106 patients receiving carboplatin reported DN compared with 75% of 47 patients treated with cisplatin and 83% of 169 patients treated with doxorubicin. Differences in the frequency of DN between doxorubicin and cisplatin were not significant (P = 0.21). Patients treated with doxorubicin were also somewhat more likely to characterize their DN as moderate or severe although greater than half of patients receiving any one of these three chemotherapy agents rated their DN as moderate or severe. Mean peak nausea severity for patients reporting DN was 4.34 for patients treated with doxorubicin, which is significantly higher than the mean peak nausea severity (3.66) for patients receiving carboplatin (P = 0.01) but not significantly different from the value (4.26) for patients treated with cisplatin (P = 0.8). Peak DN severity was comparable for patients receiving either platinum-based agent (P = 0.11).

Regression analysis confirmed these findings and indicated that gender predicted 2.9% of the variance when entered first in a hierarchical regression equation predicting peak DN severity (P = 0.002). Age, when entered next, accounted for an additional 2.6% of the variance (P = 0.003). Types of chemotherapy received, when entered at the next step using 2 orthogonal dummy-coded variables, added an additional 4% of the variance (P = 0.000) and confirmed that there was no significant difference in peak DN severity between patients receiving cisplatin and doxorubicin. Additional regression analyses using average DN severity as well as average and peak acute nausea severity as dependent variables also showed that after controlling for age and gender, the type of drug received still makes a significant difference in the frequency and severity of both acute nausea and DN.

Acute and Delayed Emesis by Chemotherapy Agent

Doxorubicin was more likely than carboplatin to cause emesis on the day of treatment (P < 0.001). There were no significant differences in the overall frequency of emesis or in the frequency of delayed emesis at Cycle 1 among the three chemotherapy agents (all Ps > 0.05). Differences in emesis frequency were more pronounced at Cycle 2 with cisplatin more likely to cause both overall emesis and delayed emesis than doxorubicin or carboplatin (Ps < 0.01). There were no significant differences by chemotherapy drug in the occurrence of acute emesis (all Ps > 0.05). No consistent differences in the severity of peak emesis by chemotherapy agent were identified at either cycle. Table 2 shows the frequency and severity of acute nausea and DN and emesis by chemotherapy agent for the 322 patients providing evaluable data at Cycle 1 and for the 300 patients at Cycle 2.

Table 2. Frequency and Severity of Acute and Delayed Nausea and Delayed Emesis by Chemotherapy Agent
CharacteristicsCarboplatin (%)Cisplatin (%)Doxorubicin (%)
Cycle 1(n = 106)(n = 47)(n = 169)
 Acute (Day 1 at its worst)  
  No nausea75.557.445.0
  Mild nausea14.221.313.6
  Moderate nausea7.517.022.5
  Severe nausea2.84.318.9
 Delayed (Days 2–5 at their worst)  
  No nausea44.325.517.2
  Mild nausea20.817.016.6
  Moderate nausea24.538.340.8
  Severe nausea10.419.125.4
Cycle 2(n = 97)(n = 43)(n = 160)
 Acute (Day 1 at its worst)  
  No nausea77.372.155.0
  Mild nausea7.211.615.6
  Moderate nausea14.411.619.4
  Severe nausea1.04.710.0
 Delayed (Days 2–5 at their worst)  
  No nausea46.425.625.0
  Mild nausea13.49.318.1
  Moderate nausea26.844.241.9
  Severe nausea13.420.915.0

Time Course of Nausea Development

Table 3 shows the day nausea first occurred and the severity and persistence of nausea that developed after the first chemotherapy cycle. Nausea was most likely to develop within the first 48 hours after administration of chemotherapy in patients receiving any of these three agents. However, 18% of the patients in this sample did not experience nausea until the third day of the cycle or later. In general, nausea that first occurred on the day of chemotherapy was more severe than nausea that first occurred in the delayed phase. Once it occurs, nausea persists. Two-thirds to three-fourths of patients who experienced this side effect on one of the first 4 days of the cycle reported that their nausea was also experienced on at least the following day.

Table 3. Persistence of Nausea after the First Occurrence of Symptomsa
Day that nausea first occurredFrequency (%)bSeverityc (SE)Persisting nausead (%)
  • a

    n = 322 (23.3% of patients reported no nausea.

  • b

    Proportion of patients experiencing nausea.

  • c

    Severity of nausea at its worst in patients reporting symptoms.

  • d

    Proportion of patients reporting nausea that continued into the following day.

Day of treatment43.23.9 (0.16)75.5
Day after treatment15.53.0 (0.19)76.0
Second day after treatment9.63.3 (0.27)74.2
Third day after treatment6.23.1 (0.26)60.0
Fourth day after treatment2.23.6 (0.90)


These results show that, despite prophylaxis on Day 1 of chemotherapy with the most effective antiemetic medications available, the majority of cancer patients receiving one of three commonly administered chemotherapy agents continued to experience nausea after chemotherapy. Of particular concern, given the lack of effective treatment modalities, is the high prevalence of DN. The frequency of DN was nearly twice as high as that of acute nausea. Once a patient experienced nausea, the nausea was very likely to persist at least through the following day. Other patients experienced nausea on one day and then had a day or more without nausea followed by recurrence of the symptom.

The pattern of nausea differed somewhat by chemotherapy agent. Doxorubicin was the most likely of the three drugs to be associated with nausea, both on the day of treatment and during the delayed period. The nausea that occurred on the day of treatment was, on average, more severe than that caused by either of the other two agents. Emesis, although less frequent than nausea, developed in approximately 25–33% of the sample, depending on the chemotherapy agent. Female gender and younger age both contributed to the variance in acute nausea and DN severity, but the chemotherapy drug remained a significant predictor. Gender, age, and chemotherapy agent together accounted for 9.5% of the variance in peak DN severity.

A recent report from the Italian Group for Antiemetic Research10 found that either dexamethasone alone or dexamethasone in combination with ondansetron effectively prevented both emesis and moderate to severe nausea compared with placebo during the delayed period in patients receiving moderately emetogenic chemotherapy (including doxorubicin and carboplatin) who did not experience these side effects on the day of chemotherapy (low-risk group). However, neither regimen effectively prevented these side effects on Days 2–5 in patients who had experienced them on the treatment day (high-risk group).10 The results of the Italian Group study as well as those from the current study confirm conclusions previously reached by our group.11 These results suggest that methods in general use by community oncologists for control of nausea and emesis caused by chemotherapy, including the use of 5HT3 receptor antagonists, were not optimal, particularly against the side effects experienced by patients at home after the day of outpatient chemotherapy administration.

A limitation of this study is that our study protocol was developed before guidelines for control of chemotherapy-induced nausea and emesis were published. Although patients were given antiemetics as prescribed by the treating oncologist during the delayed period, no specific regimen for DN, such as the use of dexamethasone, with or without a 5-HT3 receptor antagonist or metoclopramide, was mandated. Just before the beginning of accrual and during the course of this study, guidelines for the prevention of delayed emesis were published by the National Comprehensive Cancer Network,12 the Multinational Association of Supportive Care in Cancer,13 and the American Society of Clinical Oncology.14 The evidence and degree of consensus for these guidelines were moderate at best and they were considered to be voluntary for practicing physicians, who were encouraged to use their clinical judgment in each case.14 Unfortunately, because we did not assess the amount and type of antiemetic medications given during the delayed phase of this study, we cannot say how closely they approximated current guidelines. We believe that the frequency and severity of DN would be lower than we have reported if physicians follow current recommended practices.

Another limitation of the study is inherent to all studies relying on self-report measures completed by patients at home. It is possible, considering that approximately 10% of patients in the current study did not provide evaluable data, that our results are inflated by a report bias favoring patients who experienced nausea.

Despite the acknowledged limitations of this study, it provides valuable information concerning the high prevalence of DN, especially that associated with doxorubicin, that is important to today's practitioners. This is particularly relevant in light of reports published in the medical literature regarding prescribing practices of community oncologists, which indicate that physicians often do not follow clinical practice guidelines developed by professional groups. In a 1998 report by the Italian Group for Antiemetic Research,15 only 33% of 742 patients treated with moderately emetogenic chemotherapy in centers in which physicians had previously participated in antiemetic research studies received prophylaxis for DN. In addition, only 27% were given the then-preferred treatment, a serotonin receptor antagonist or a steroid, alone or in combination. More recent reports confirm that the trend for physician noncompliance with antiemetic guidelines for DN continues. In two reports, approximately one-fourth of patients being studied were prescribed postchemotherapy steroids for the prevention of delayed chemotherapy-induced nausea and emesis.16, 17

Poor compliance with guidelines may be related to the underestimation of the frequency of chemotherapy-related DN by physicians and nurses.1718 It is also possible that physicians do not have the resources or the time to manage antiemetic use beyond the clinic or that they are not persuaded that the guidelines present the most effective regimens for handling DN.

Reports that show that nausea is not well controlled in normal clinical practice and educational efforts that emphasize the significant adverse effects of this side effect on patients' quality of life and ability to perform everyday tasks may improve physician compliance with published guidelines.17 Additional research studies that emphasize DN and employ recommended agents as outlined in current antiemetic guidelines are also warranted as they will help to strengthen physicians' acceptance. These should take into account and build on the finding that dexamethasone reduces the occurrence of DN, suggesting that the hypothalamic-pituitary-adrenal (HPA) axis may be involved in its development.10 Ultimately, better control of DN may depend on further elucidation of the causal mechanism of this common side effect of chemotherapy.


The authors thank the following Community Clinical Oncology Program (CCOP) Principal Investigators (PI) who provided and cared for study patients as they collected data: Paul Weiden, M.D., former PI, Virginia Mason Research Center CCOP, Seattle, WA; Philip J. Kuebler, M.D., Columbus CCOP, Columbus, OH; Brian Issel, M.D., Hawaii MBCCOP, Honolulu, HI; Tarit K. Banerjee, M.D., Marshfield Medical Research Foundation CCOP, Marshfield, WI; Patrick J. Flynn, M.D., Metro-Minnesota CCOP, St. Louis Park, MN; Richard J. Rosenbluth, M.D., Northern New Jersey CCOP, Hackensack, NJ; H. Irving Pierce, M.D., former PI (deceased), Northwest CCOP, Tacoma, WA; and Jeffrey J. Kirshner, M.D., Syracuse Hematology-Oncology CCOP, Syracuse, NY. The authors also thank Dr. Richard Gralla for his thoughtful review of the article. Many of his valuable suggestions have been incorporated into the article. Shonda Ranson is acknowledged for her outstanding work in assuring the accuracy and completeness of the database.