2024 Volume 73 Issue 2 Pages 15-23
Radiation-induced nausea and vomiting (RINV) is a frequent adverse event that occurs in patients undergoing radiotherapy. However, research on RINV is underrepresented. This prospective single-institution exploratory pilot study investigated the incidence of RINV according to the irradiation site and observed the efficacy of symptomatic antiemetic treatment in controlling symptoms of RINV. The primary outcomes were the proportions of emesis-free days and nausea-free days. The secondary endpoints included the time to the first episode of RINV, frequency of vomiting, and severity of nausea, including its impact on eating habits and weight loss. Fifteen patients were enrolled in each group (minimal, low, and moderate emetogenic risk). All patients received greater than 20 Gy in five fractions. Evaluation was based on weekly questionnaires completed by patients during routine clinic visits. Nausea and vomiting occurred in 11 and 0 patients, respectively. Six of 15 patients in the minimal-risk group, 1 in the low-risk group, and 4 in the moderate-risk group experienced nausea. Although all 11 symptomatic patients were offered antiemetics, only 3 used them, who reported satisfactory control of nausea. The percentage of emesis-free days for all patients was 100% and the percentage of nausea-free days for the 11 patients who developed RINV was 38%. An unexpectedly high percentage of patients in the minimal-risk group experienced nausea; all had breast cancer. Future studies should investigate factors beyond the irradiation site, including the characteristics of the patient and the treatment, to better predict an individual’s risk of RINV.
Approximately 50% of all cancer patients receive radiotherapy during their course of the disease, with palliative or curative intent.1 Radiation-induced nausea and vomiting (RINV) is a frequent and distressing adverse event (AE) that occurs in 50%–80% of patients undergoing radiotherapy.2,3,4 RINV can have a negative impact on patient health-related quality of life (HRQoL), resulting in emotional distress, inability to perform daily activities, tiredness, loss of appetite, weakness, and disruption of social and work life.4,5,6 In severe cases, uncontrolled RINV can lead to potentially life-threatening complications, such as dehydration and electrolyte imbalances.7,8 RINV can require hospitalization, lead to treatment delays or termination of radiotherapy, and compromise treatment outcomes.7,8 Nevertheless, research on RINV is underrepresented relative to research on chemotherapy-induced nausea and vomiting.7
Several factors are believed to influence the onset of RINV, including age, sex, irradiation site, treatment field size, and dose fractionation.2,9 Among them, evidence supporting the irradiation site as a key factor has been supported by several clinical studies and expert opinions; it has been used to generate a risk classification system in antiemetic guidelines published by the Multinational Association of Supportive Care in Cancer (MASCC)/European Society for Medical Oncology (ESMO), American Society of Clinical Oncology (ASCO) and National Comprehensive Cancer Network (NCCN).1,9,10,11,12 This classification defines emetogenic risk as high (>90%), moderate (30%–90%), low (10%–30%), or minimal (<10%), depending on the irradiated site. However, awareness of antiemetic guidelines remains limited among radiation oncologists, and clinical practice patterns vary.3 Radiation oncologists may underestimate the risk of RINV and neglect to ask patients about nausea and vomiting; as a result, symptoms may go undetected, and patients may be undertreated.2
Based on our clinical experience, we did not recognize any prominent differences in the incidence of RINV according to the irradiated site. However, given that it is not customary to ask patients for in-depth details about nausea and vomiting, it is possible that we were unaware of the onset of RINV.
We conducted a prospective exploratory pilot study of patients from our hospital who received radiotherapy alone to observe the incidence of RINV according to the risk group and to observe the efficacy of symptomatic antiemetic treatment in controlling the symptoms of RINV. Our study addresses biases observed with clinician-reported toxicity ratings by using patient-reported outcomes (PROs).13
From March 2020 to October 2021, 47 consecutive patients who met the eligibility criteria were prospectively recruited at our hospital. Forty-five patients were enrolled in the study: 15 patients for each of the minimal-, low-, and moderate-emetic-risk groups. We used the common risk classification used by MASCC/ESMO, ASCO, and NCCN that defines the emetogenic risk associated with various radiation sites as minimal (breast, extremities), low (brain, head and neck, thorax, pelvis), and moderate (upper abdomen, cranial–spinal). When 15 patients were enrolled and completed the intended therapy in each risk group, that group was closed. We chose not to include high-risk patients (total body irradiation) because total body irradiation is performed rarely in our hospital, and it would be difficult to discern the effects of radiotherapy alone because patients routinely receive emetogenic chemotherapy.
The following inclusion criteria were used: age over 20 years, receiving a photon radiotherapy dose greater than 20 Gy in five fractions, and capable of responding to weekly questionnaires. Patients were excluded for the following reasons: receiving concomitant chemotherapy, undergoing stoma construction, being diagnosed with ileus, or other reasons judged as inappropriate by the attending physician for participation in the trial. Patients receiving concomitant chemotherapy were excluded because they usually receive antiemetics, and we wanted to avoid the confounding emetogenic effect of chemotherapy. Given that we wanted to mimic the ordinary clinical setting, prophylactic antiemetics were not administered.
Based on patient responses to a questionnaire, nausea was assessed at the time of enrollment, at weekly intervals during the radiotherapy period, and a month after the completion of radiotherapy. The questionnaire was based on the MASCC Antiemesis Tool and was modified for this study (Appendix A). It included questions about episodes of vomiting, prevalence of nausea, the degree of nausea based on a visual analog scale ranging from 0 to 10, and effects on daily eating habits. Acute AEs, including nausea and vomiting, were classified according to the National Cancer Institute Common Terminology Criteria for AEs v5.0.
Patients diagnosed with RINV were offered granisetron hydrochloride [an oral 5HT3-receptor antagonist (5HT3-RA)], which was administered as one dose (two 1-mg tablets) 1 h prior to the start of radiotherapy. For patients with very mild nausea, the dose was reduced to 1 mg at the discretion of the physician.
The primary endpoint of the study was the proportion of emesis-free days and nausea-free days, as defined by the number of days within the radiotherapy period. Secondary endpoints included the time to the first episode of RINV, the number of times that vomiting occurred, and the severity of nausea, including the effect on eating habits and weight loss.
Ethical approvalAll procedures were in accordance with the ethical standards of the responsible committee at which the studies were conducted (IRB approval number 20190181) and with the Declaration of Helsinki of 1964 and later versions. The trial was registered with the University Hospital Medical Information Network (UMIN000046427). Written informed consent for inclusion in the study and publication of data was obtained from all patients prior to the first day of radiotherapy.
Written informed consent was obtained from 47 patients; 1 patient experienced cerebral hemorrhage after receiving 24 Gy of radiotherapy, and 1 patient died before receiving any therapy. Therefore, 45 patients who completed the intended therapy were included. The patient characteristics and radiotherapy characteristics are summarized in Tables 1 and 2, respectively.
| Characteristics | n | |
|---|---|---|
| All | 45 | |
| Age (years) | Median (range) | 65 (41–87) |
| <65/≥65 years | 24/21 | |
| Sex | Male/female | 16/29 |
| History of alcohol use | Yes/no | 27/18 |
| History of tobacco use | Yes/no | 25/20 |
| ECOG PS | 0–1/2–3 | 37/8 |
| Previous chemotherapy | Yes/no | 26/19 |
| Risk classification and site | Minimal/Low/Moderate | 15/15/15 |
| Minimal: Breast/Extremities | 14/1 | |
| Low: Brain/Head and neck/Thorax/Pelvis | 1/3/8/3 | |
| Moderate: Upper abdomen | 15 | |
| Tumor status of radiotherapy site | Primary/metastasis | 32/13 |
| Prescribed dose (Gy) | ≤40/>40 | 21/24 |
| Irradiated volume (cm3) | ≤400/>400 | 33/12 |
ECOG, Eastern Cooperative Oncology Group; PS, performance status.
| Location | n | Prescribed dose (Gy) median (range) |
Fractions median (range) |
Irradiated volume (cm3) mean (range) |
RT technique (3D-CRT/IMRT) |
|---|---|---|---|---|---|
| Minimal | 15 | ||||
| Breast | 14 | 50 (43–60) | 25 (16–30) | 394 (196–1169) | 14/0 |
| Extremities | 1 | 66 (N/A) | 33 (N/A) | 1493 (N/A) | 1/0 |
| Low | 15 | ||||
| Brain | 1 | 20 (N/A) | 5 (N/A) | 121 (N/A) | 1/0 |
| Head and neck | 3 | 60 (20–66) | 30 (5–33) | 158 (27–383) | 0/3 |
| Thorax | 8 | 42 (20–66) | 20 (5–33) | 248 (84–175) | 5/3 |
| Pelvis | 3 | 20 (20–30) | 5 (5–10) | 215 (175–278) | 3/0 |
| Moderate | 15 | ||||
| Upper abdomen | 15 | 30 (20–66) | 10 (5–33) | 448 (65–1018) | 11/4 |
RT, radiotherapy; 3D-CRT, three-dimensional conformal radiotherapy; IMRT, intensity modulated radiotherapy; N/A, not applicable.
None of the patients reported nausea or vomiting at the time of enrollment. Likewise, none of the patients had undergone chemotherapy within 1 week of starting radiotherapy. Nausea and vomiting occurred in 11 of 45 (24%) and 0 of 45 (0%) patients, respectively. None of the patients answered “yes” to question 1: “In the previous week of radiotherapy, did you have any vomiting?” Therefore, all patients skipped question 2: “If you vomited in the previous week of radiotherapy, how many times did it happen?” The 11 patients who answered “yes” to question 3: “In the previous week of radiotherapy, did you have any nausea?” were questioned further and their medical records were scrutinized to exclude causes of nausea other than radiotherapy. Some patients had received opioids, which are known for their emetogenic potential. However, these patients had not reported nausea at any other time during the study period. All 11 patients were considered to have developed RINV, and their characteristics are summarized in Table 3. Patients’ responses to question 4: “How much nausea did you have in the last week?” were graded according to VAS and are summarized in Table 3 and Fig. 1. Most patients scored their nausea as mild with a maximum score of less than 3. Six of the 15 patients (40%) in the minimal-risk group, 1 patient (7%) in the low-risk group, and 4 patients (27%) in the moderate-risk group experienced nausea. The 6 patients in the minimal-risk group had breast cancer. Patients who developed nausea and answered “2” to question 5: “Did nausea affect your eating habits?” were graded as having grade 2 nausea, whereas patients answering “1” were graded as having grade 1 nausea. Among the 11 patients who experienced nausea, 4 patients had grade 1 nausea and 7 patients had grade 2 nausea. None of the patients developed grade 3 nausea.
| Patient no. | Disease (status) | RC | Sex | Age (years) | Pres. dose (Gy |
Frac. | Irrad. vol. (cm3) |
Nausea gradea |
Visual analog scaleb | Medication | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Peak score |
Week no. | ||||||||||
| 1 | Breast ca. (prim) | Min | F | 45 | 50 | 25 | 272.6 | 1 | 2 | 1, 2, 3, 4, 5 | Yes |
| 2 | Breast ca. (prim) | Min | F | 49 | 50 | 25 | 592.7 | 1 | 3 | 1, 2 | No |
| 3 | Breast ca. (prim) | Min | F | 50 | 60 | 30 | 318.5 | 2 | 3 | 1 | No |
| 4 | Breast ca. (prim) | Min | F | 42 | 52.6 | 21 | 399.8 | 1 | 2 | 1 | No |
| 5 | Breast ca. (prim) | Min | F | 41 | 60 | 30 | 1169.0 | 1 | 3 | 4 | No |
| 6 | Breast ca. (prim) | Min | F | 52 | 50 | 25 | 216.7 | 2 | 8 | 1 | No |
| 7 | PRMS (neck tumor bed) | Low | M | 75 | 60 | 30 | 383.2 | 2 | 2 | 3 | No |
| 8 | Esophageal ca. (lumbar meta) | Mod | F | 57 | 20 | 5 | 334.2 | 2 | 2 | 1 | No |
| 9 | Gastric MALT lymphoma (prim) | Mod | M | 70 | 30 | 15 | 861.9 | 2 | 1 | 1, 3 | No |
| 10 | Lung ca. (lumbar meta) | Mod | F | 56 | 30 | 10 | 243.6 | 2 | 7 | 1 | Yes |
| 11 | Sigmoid colon ca. (lumbar meta) | Mod | F | 62 | 30 | 10 | 173.6 | 2 | 6 | 2 | Yes |
RINV, radiotherapy-induced nausea and vomiting; ca., cancer; prim, primary; meta, metastasis; MALT, mucosa-associated lymphoid tissue; PRMS, pleomorphic rhabdomyosarcoma; RC, risk classification; Min, minimal; Mod, moderate; F, female; M, male; Pres., prescribed; Frac., fractions; Irrad. vol., irradiated volume.
a Nausea grade according to the National Cancer Institute Common Terminology Criteria for Adverse Events v.5.0.
b Visual analog scale: patient-rated scale from 0 to 10.
Change in severity of nausea over time for 11 patients that developed RINV.
(A) Minimal-emetic-risk group (6 patients), (B) low-emetic-risk group (1 patient), and (C) moderate-emetic-risk group (4 patients). MALT, mucosa-associated lymphoid tissue.
The time from the start of radiotherapy to the onset of nausea was within 1 week for 9 patients and within 2 weeks for 10 patients (Fig. 1). Nausea generally persisted throughout the course of radiotherapy and was alleviated within 2 weeks after completion. No patient reported nausea at the 1-month follow-up visit. All symptomatic patients were offered antiemetic therapy with oral 5HT3-RA; however, only 3 patients (27%) took the medication. The percentage of emesis-free days for all patients was 100%. The percentage of nausea-free days for the 11 patients who developed RINV was 38%. For the 3 patients who took the medication, the percentage of nausea-free days was 26% (range 0%–79%), and for the 8 patients who did not, it was 45% (range 0%–97%). Patients receiving the medication reported satisfactory control of nausea, but none achieved complete control.
AEs in the study cohort are summarized in Table 4. Eleven (24%) patients experienced nausea; 19 (42%) patients had radiation dermatitis, including 3 patients with grade 3 dermatitis; and 14 (31%) patients experienced bone pain. None of the patients discontinued treatment because of AEs.
| Adverse event | Any grade | Grade 1 | Grade 2 | Grade 3 |
|---|---|---|---|---|
| Nausea | 11 | 4 | 7 | 0 |
| Radiation dermatitis | 19 | 6 | 10 | 3 |
| Alopecia | 1 | 0 | 1 | 0 |
| Diarrhea | 2 | 1 | 1 | 0 |
| Urinary frequency | 2 | 2 | 0 | 0 |
| Constipation | 1 | 1 | 0 | 0 |
| Bone pain | 14 | 2 | 12 | 0 |
| Esophagitis | 1 | 1 | 0 | 0 |
| Peripheral sensory neuropathy | 3 | 3 | 0 | 0 |
| Edema | 1 | 1 | 0 | 0 |
After observing that an unexpectedly large percentage of patients with breast cancer developed RINV (6 of 14 patients, 43%), we decided to further analyze their backgrounds to determine if any factors were relevant (Table 5). The following patients had a greater tendency to develop RINV: those younger than 65 years, those with a history of alcohol use, those with a history of tobacco use, those without previous chemotherapy, and those with a radiation volume greater than 400 cm3.
| Characteristic | All cases n |
Nausea n (%) |
No nausea n (%) |
|
|---|---|---|---|---|
| All patients with breast cancer | 14 | 6 (43) | 8 (57) | |
| Age (years) | <65 | 10 | 6 (60) | 4 (40) |
| ≥65 | 4 | 0 (0) | 4 (100) | |
| History of alcohol use | Yes | 8 | 4 (50) | 4 (50) |
| No | 6 | 2 (33) | 4 (67) | |
| History of tobacco use | Yes | 5 | 3 (60) | 2 (40) |
| No | 9 | 3 (33) | 6 (67) | |
| ECOG PS | 0 | 9 | 4 (44) | 5 (56) |
| 1 | 5 | 2 (40) | 3 (60) | |
| Previous chemotherapy | Yes | 8 | 3 (38) | 5 (63) |
| No | 6 | 3 (50) | 3 (50) | |
| Prescribed dose (Gy) | ≤40 | 0 | 0 (0) | 0 (0) |
| >40 | 14 | 6 (43) | 8 (57) | |
| Irradiated volume (cm3) | ≤400 | 9 | 3 (33) | 6 (67) |
| >400 | 5 | 3 (60) | 2 (40) | |
ECOG, Eastern Cooperative Oncology Group; PS, performance status.
In our study, 24% of patients experienced nausea and no patients experienced vomiting during the study period. Both incidence rates were lower than those in previous reports. In a landmark study on RINV, conducted by the Italian Group for Antiemetic Research in Radiotherapy, the cumulative total incidence of vomiting and nausea was approximately 28%, and nausea was twice as common as vomiting (27.1% vs 11.0%). However, when limited to patients receiving radiotherapy alone, the incidence was 23.7%.2 In a study from Sweden, during a typical week of treatment, 39% of patients experienced nausea and 7% vomited; however, the percentage of patients who received radiotherapy alone was unclear in that study.6 The lower incidence of RINV in our study may be attributed to modern radiotherapy techniques, which allow high conformality and small radiation fields. In addition, the small sample size and exclusion of patients receiving concomitant chemotherapy in our study may have contributed to this discrepancy.
Radiotherapy to the upper abdomen has been reported as the most important risk factor for RINV. Furthermore, 50% of patients treated in the upper abdomen in the Italian trial2 and 63% of patients treated in the abdominal or pelvic fields in the Swedish study6 experienced RINV. In our study, only 27% of the patients in the moderate-risk group (receiving abdominal radiotherapy) experienced nausea. In contrast, we observed an unexpectedly high incidence of RINV in the minimal-risk group (40%), of whom, 14 of 15 patients had breast cancer. When compared with the results of the Japanese JCOG 0906 trial of hypofractionated whole breast radiotherapy,14 the incidence of nausea observed in the current study was significantly higher. However, the results are difficult to compare because our study included only a few hypofractionated patients (3 of 14), whereas all patients in the JCOG 0906 trial received hypofractionated radiation. Furthermore, the JCOG 0906 trial used physician-reported toxicities, whereas our study relied on PROs. We believe that the difference in the toxicity evaluation method could be the main reason for the higher incidence of nausea in our study. Data regarding minimal-risk patients are lacking, and guideline recommendations are inconsistent.2,3,4
The risk classification for radiotherapy has been defined only by the location of radiotherapy despite the influence of many other factors on the incidence and severity of RINV.3,4,15 Currently, the only accepted individual-level risk factor for RINV is previous or concomitant chemotherapy, although various other potential patient-related factors have been identified, such as a history of alcohol consumption, sex, age, and previous episodes of nausea and vomiting.2,9 In our study, breast cancer patients with a history of alcohol or tobacco use showed a higher tendency to develop RINV, which is in line with previous reports. Conversely, those who had not been treated with chemotherapy before radiotherapy exhibited a higher tendency to develop RINV than those who had received chemotherapy. This observation is contrary to common knowledge; nearly 25% of chemotherapy recipients in a prominent PRO study reported severe/very severe nausea/vomiting during their cancer treatment.13 However, our observation is similar to data from groups exploring radiotherapy and fatigue: patients with previous chemotherapy reportedly experienced less fatigue than those without prior chemotherapy.16,17 Several studies have demonstrated high rates of burdensome toxicities reported by women with early-stage breast cancer, supporting recent assertions that many women with curable disease experience “collateral damage” from breast cancer treatment.13,18 Some studies have found that RINV survivors report a number of symptoms and these vary according to the treatment received.19 Patients with RINV also tended to be younger, which is consistent with other studies wherein older patients were more likely to report moderate to good quality of life.20,21 Some investigators have hypothesized that older patients are mentally better equipped to deal with treatment compared with younger patients.20,22 How RINV onset correlates with general HRQoL of patients is a topic that warrants further investigation. We speculate that because radiotherapy is performed as a preventive approach for breast cancer patients who have a stable baseline performance status, nausea caused by radiotherapy may have a higher incidence and a more profound effect on HRQoL than that in patients with other types of cancer. Conversely, although we observed a higher incidence of nausea in breast cancer patients than in other studied groups, the degree to which food intake was affected was less than that in other groups: four of the six breast cancer patients that developed nausea reported grade 1 nausea, whereas all five patients in the low- and moderate-risk groups reported grade 2 nausea. It is possible that the perception of nausea by breast cancer patients is greater than that of other cancer patients, which is not correlated with physical aspects. Specific to breast cancer, the experience of diagnosis and treatment of breast cancer can cause considerable distress, and we speculate that this background may also contribute to the higher incidence of RINV in breast cancer patients.22 Patients may also experience anticipatory nausea. This response is similar to anticipatory emesis associated with chemotherapy, which occurs predominantly in women, although its occurrence in response to radiotherapy remains unclear.
Treatment-related factors, such as total radiation dose, fractionation schedule, field size, and delivery technique, can also affect the onset of RINV.9 Although the radiotherapy technique was not described in detail in their study, Maranzano et al.2 reported that the irradiation site and field size greater than 400 cm2 were treatment-related factors that were significantly associated with RINV, suggesting that radiotherapy was applied via two-dimensional portals. Modern radiotherapy techniques, such as three-dimensional conformal radiotherapy and intensity-modulated radiotherapy, allow highly conformal, smaller treatment volumes, which can be calculated in cubic centimeters. Our results support the observation that patients receiving radiotherapy with a larger field size have a higher probability of RINV; however, we were able to use a cutoff value of 400 cm3 (cm3 rather than cm2), which is a much lower value than that previously reported (>400 cm2). As stated above, smaller treatment volumes likely accounted for the low incidence of RINV in the low- and moderate-risk groups in our study. Therefore, we discovered a more pronounced incidence of RINV among patients with breast cancer than among patients with other types of cancer, highlighting the possibility that such patients should not be classified as being at minimal risk, at least in Japan. Some groups have questioned whether a racial/ethnic discrepancy exists in the patient threshold for reporting symptoms or whether they truly have a different degree of symptoms.19,23
Although clinical guidelines recommend prophylactic antiemetic use for patients undergoing radiotherapy at sites of more than moderate risk of RINV, we chose to provide only rescue therapy for all risk groups to reflect common clinical practice. The 5HT3-RA granisetron hydrochloride was selected because it is the principal constituent of RINV antiemetic regimens and is one of the most well-studied antiemetics for radiotherapy, with a proven record of success.24,25 However, patients included in most antiemetic studies received total body irradiation or radiotherapy to the upper abdomen, and evidence concerning nausea prevention rather than control of vomiting secondary to radiation is lacking, especially for the low- and minimal-risk groups.4,7,25,26 In our study, the primary endpoint of emesis-free days was 100%, suggesting that prophylactic administration of antiemetic medication is not needed, at least for the minimal-, low-, and moderate-emetic-risk groups. The three patients who chose to use antiemetic medication reported satisfactory nausea control, although treatment was ineffective in achieving complete control of nausea as shown by patient VAS scores. Notably, whether a patient chose to take medication did not directly correlate with the severity of nausea. The evaluation of efficacy by nausea-free days was not feasible because of the few patients taking medication and the potential interference of recall bias when patients completed the weekly questionnaires; however, symptomatic administration after careful patient observation appears to be sufficient to manage nausea for the risk groups included in our study.
Our study has some limitations. First, this was an observational study, and we did not control for patient characteristics, such as radiation field size, sex, or age. Patients also received a variety of radiation doses, because patients with both definitive and palliative intent radiotherapy were included. The rationale for including patients with more than 20 Gy in five fractions was based on data from a previous study reporting that patients with RINV generally experienced nausea within 3 days after starting radiotherapy.2 However, limiting patients to those receiving either palliative or definitive radiotherapy may have impacted on the results, not only because of discrepancies in radiation dose but also patient characteristics such as performance status and metastatic state. The study population was small and the inclusion of a larger number of patients may also have led to different results. However, although our study was underpowered to perform a statistical analysis of emetogenic risk factors, our observations suggest that the emetogenic risk for the minimal-risk group, especially for patients with breast cancer, could be underestimated.
In conclusion, although the overall incidence of RINV in our hospital was lower than that in previous reports, the incidence in patients with breast cancer was higher than expected. The reasons for the high incidence of RINV in patients with breast cancer should be further explored, including the factors unique to this group. Given that the irradiation site alone is insufficient to determine an individual’s risk of RINV, other patient, treatment, and dosimetric factors should be studied.
We thank Editage (www.editage.com) for English language editing.
The authors declare no conflict of interest.
