Introduction

Past observational studies on radiotherapy-induced nausea and vomiting (RINV) highlight that the overall cumulative incidence of vomiting and nausea is about 50–80% of patients undergoing radiotherapy [10, 11, 29]. It was also highlighted that the attitude of radiation oncologists in prescribing antiemetic drugs as a rescue, with a large range of doses and schedules, and that 5-hydroxytryptamine (5-HT3) receptor antagonists rather than other antiemetics were generally being used [42, 7, 9, 14, 29]. In 2009, undertreatment of RINV of patients undergoing radiotherapy is still a reality, as shown in two large observational trials [5, 29].

Patients submitted to total body irradiation (TBI), half body irradiation (HBI) or abdominal radiotherapy are at a major risk of nausea and vomiting. Poorly controlled nausea and vomiting can result in potentially life-threatening medical conditions, including dehydration and electrolyte imbalance. Some patients find nausea and vomiting so distressing that they are willing to compromise their treatment regimens to avoid these symptoms. As such, this may have serious consequences for treatment success. Studies have revealed an increased risk of treatment failure and poor tumour control, when for example, the overall radiotherapy time is increased. Few randomized controlled clinical trials have evaluated the efficacy of various antiemetic drugs in preventing RINV. Since the last update in 2005 [7, 9], only one well-designed double-blind randomized study in patients undergoing upper abdominal radiation treated with a 5HT3 receptor antagonist in combination with short-course dexamethasone versus a 5-HT3 receptor antagonist alone was published [46].

The purpose of this updated MASCC/ESMO guideline is to provide a consensus statement derived from published articles as well as expert opinion about antiemetic therapy for radiotherapy-induced nausea and vomiting. This specific recommendation does not include a guideline on radiotherapy-induced nausea and vomiting in children.

Literature search strategy

A Medline search from January 2004 to June 2009 was conducted for English language articles using the search terms: “radiotherapy-induced emesis” and radiotherapy-induced nausea”. Abstracts were reviewed, and articles were excluded if they possessed any of the following characteristics: review articles, cause for emesis other than radiotherapy or radiochemotherapy.

Current practice guidelines for RINV

Current Multinational Association of Supportive Care in Cancer (MASCC), American Society of Clinical Oncology (ASCO) and National Cancer Comprehensive Network (NCCN) practice guidelines for the use of antiemetics in radiotherapy are in detail different when classifying radiation emetogenic risk categories and the indications for the use of antiemetic drugs [7, 9, 22, 33, 45]. This diversity of recommendations reflects the limited amount of high-level evidence available to date (i.e. few randomized controlled trials and the small number of patients entered in each trial).

Observational trials

The incidence and management of RINV have been recently updated in two prospective observational studies [5, 29]. The first survey showed that although approximately one third of radiotherapy patients experienced nausea and vomiting, the vast majority (85%) were not prescribed antiemetics [5].

In the latest observational study by the Italian Group for Antiemetic Research in Radiotherapy (IGARR), 1,020 patients receiving different kinds of radiotherapy were included [29]. Overall vomiting and/or nausea was reported by 280 (27.9%) of the patients. The median time to the first episode of vomiting was 3 days. Antiemetic drugs were administered to 170 patients (17%) of the 1,004 evaluable patients; whereas 12.4% received prophylactic antiemetic treatment and 4.6% rescue treatment. These results are in accordance with the previous study (IGARR) and reinforce that the tendency of radiation oncologists is not to prescribe antiemetics. Additionally, in the previous study, only a minority (14%) of patients received an antiemetic drug, and the prescriptions were more often symptomatic rather than prophylactic (9% and 5%, respectively) [42].

A European survey on 200 radiation oncologists from France, Italy, Germany, Spain and the UK suggested that 5-HT3 receptor antagonists are underused in patients receiving radiotherapy. Only 52% of patients who received highly or moderately emetogenic radiotherapy actually received a 5-HT3 receptor antagonist. There are also differences in the prescribing procedure among the evaluated countries [8]. A 5-HT3 receptor antagonist was more frequently prescribed if the patient received radiation with chemotherapy than in radiotherapy alone (46% versus 33%).

Risk classification

One of the main stumbling blocks to effective treatment has been a lack of consensus on the potential for emesis with different radiotherapy techniques and doses. The updated review of the literature reveals only two suitable new references, derived from two large prospective observational studies [5, 29]. In the second IGARR study by Maranzano et al., a higher incidence for vomiting than would be expected from the current risk classification for patients receiving head and neck and brain radiation was observed. In the study by Enblom et al., the results were similar; however, nausea was defined as one of the major endpoints. Due to the results from these two studies, a reclassification of head and neck and brain radiation to the low emetogenic (last consensus: minimal risk) risk group is necessary. The risk classification is depicted in Table 3. In the second IGARR study, a multifactorial analysis evaluated the significant risk factors for RINV [29]. Concomitant radiochemotherapy, previous experience of vomiting induced by chemotherapy, irradiated site (upper abdomen) and field size (> 400 cm2) were significantly correlated with a higher incidence of RINV.

Antiemetics-randomized clinical trials

Few small randomized clinical trials have evaluated the efficacy of various antiemetic drugs in preventing RINV. Generally, patients entering these trials are those submitted to TBI, HBI or upper-abdomen irradiation because of the higher risk of developing nausea and/or vomiting.

Prophylaxis with non-5-HT3receptor antagonists

Three randomized trials on RINV in patients treated with fractionated radiotherapy to the abdomen and thorax were published before the introduction of 5-HT3 receptor antagonists. In the first study, 39 patients were randomized to receive oral metoclopramide or nabilone. In the second, 89 patients were treated with oral metoclopramide, prochlorperazine or placebo, and in the third, 11 patients received tetrahydrocannabinol or prochlorperazine [7, 9, 35, 44]. Only one randomized study has been carried out with 43 patients submitted to single-fraction palliative radiotherapy to the thoracic and/or lumbar spine. In this study, chlorpromazine was compared with two different doses of levonantradol [25].

All of these studies enrolled a small number of patients (median 46) and showed no difference among the various compounds determining a limited antiemetic efficacy (complete protection of vomiting in only about 50% of cases).

Prophylaxis with 5-HT3 receptor antagonists

In the last decade, the 5-HT3 receptor antagonists have been used more extensively in clinical practice to treat RINV [7, 9]. Tables 1 and 2 show randomized trials with 5-HT3 receptor antagonists and/or corticosteroids in patients submitted to radiotherapy with single or fractionated regimens. Different compounds and a wide range of doses and schedules have been used. One interesting trial evaluated the efficacy of an escalating dose of oral ondansetron (OND) in the prevention of emesis induced by fractionated radiotherapy. The dose-adapted regimen of OND was effective and showed the possibility to reduce costs without compromising the activity [26]. Antiemetics were generally started 1–2 h before radiotherapy and usually continued until the end of irradiation when a fractionated regimen of dose was adopted. The oral route was predominant (70%).

Table 1 Randomized clinical trials with 5-HT3 receptor antagonists and/or steroids in patients submitted to upper abdomen irradiation
Full size table
Table 2 Randomized clinical trials with 5-HT3 receptor antagonists in patients submitted to total body irradiation (TBI) and half body irradiation (HBI)
Full size table

The seven published trials mostly with ondansetron regarding patients submitted to upper-abdomen irradiation showed that 5-HT3 receptor antagonists achieved significantly greater protection for RINV than metoclopramide, phenothiazines or placebo (Table 1; [1, 3, 12, 23, 36, 37]). Also, in patients treated with TBI or HBI, 5-HT3 receptor antagonists provided as expected, significantly better protection for RINV than placebo or conventional antiemetics (Table 2; [34, 39–41, 43].

Furthermore, Spitzer [40] showed in his randomized trial that granisetron (GRAN) and OND are equally effective in controlling emesis during TBI. One further trial confirmed that a 5-HT3 receptor antagonist (GRAN) achieved a significantly better protection against RINV than placebo in patients undergoing fractionated upper-abdomen irradiation [23]. This trial compared 1 mg of oral GRAN bid versus 2 mg GRAN once a day in patients submitted to fractionated radiotherapy to abdomino-pelvic area and showed no significant differences between the two modalities in controlling RINV. Further reports support the efficacy of 5-HT3 receptor antagonists in this setting [19, 21].

The most commonly used 5-HT3 receptor antagonists are ondansetron and granisetron. Side effects were evaluated and compared by Goodin [13]. The adverse effects of 5-HT3 receptor antagonists are generally mild, with headache, constipation, diarrhea and asthenia mainly described [18, 34, 37, 40]. Sometimes, rather than causing constipation, 5-HT3 receptor antagonists reduced the frequency of diarrhea, a troublesome side effect due to acute radiation enteric toxicity [10–12].

The relatively new 5-HT3 receptor antagonist palonosetron and transdermal granisetron patch might be interesting options especially for patients receiving radiotherapy. To date, only a few abstracts are available on the use of palonosetron and the transdemal granisetron patch in RINV showing promising activity [2, 4]. Therefore, no specific recommendation for these 5-HT3 receptor antagonists is currently possible. Further research evaluating the potential role of palonosetron and the transdermal granisetron patch is warranted.

Prophylaxis with a 5-HT3receptor antagonist + dexamethasone

The effect of adding a short course of low-dose dexamethasone (days 1–5) to a 5-HT3 receptor antagonist was assessed in a National Cancer Institute of Canada (NCIC) trial, in which 211 patients receiving RT to the upper abdomen as shown in Table 1 [46]. Patients were randomly assigned to ondansetron (8 mg twice a day) plus either dexamethasone (4 mg daily) or placebo with the first five treatment fractions. Patients with persistent nausea and/or vomiting received ondansetron plus prochlorperazine with subsequent RT treatments as a rescue option. During the first 5 days, there was a non-significant trend toward improved complete control of nausea (50% versus 38% with placebo) and vomiting (78% versus 71%) (primary end point not reached). However, the effects of dexamethasone extended beyond the initial period, significantly more patients had complete control of emesis over the entire course of RT (23% versus 12%, with placebo) (secondary end point was reached). Although this study did not demonstrate a statistically significant benefit for the primary end point, results on several secondary end points, as well as quality of life data, strongly suggest that benefits do exist with the addition of dexamethasone, and therefore, dexamethasone is now recommended in combination with a 5-HT3 receptor antagonist as an option for patients in the moderate emetogenic risk category (Table 3).

Prophylaxis with corticosteroids

Their widespread availability, low cost and benefit make corticosteroids to be a very interesting antiemetic drug. Regarding the use of dexamethasone (DEX) as a single agent for the prophylaxis of RINV, a double-blind study has been published so far [20]. Patients who underwent fractionated radiotherapy to the upper abdomen received either oral DEX (2 mg × 3/day) or placebo only in the first week of radiotherapy, even though the courses lasted up to 6 weeks (Table 1). Complete protection from RINV was significantly better in the DEX group with acceptable side effects. However, no overall positive effect on global quality of life could be detected. Considering that the majority of emetic episodes occurred early in the treatment course, it is indeed arguable that prophylactic antiemetics may not be necessary for a full course of radiotherapy, and therefore, a treatment for the first week might be sufficient [7, 9, 20, 22]. More studies evaluating the efficacy of steroids in comparison to 5-HT3 receptor antagonists and as a combination regimen are further warranted to enlarge the body of evidence.

Prophylaxis with neurokinin-1 receptor antagonists

Extensive clinical trials have established an important role for the neurokinin-1 (NK-1) receptor antagonists in the management of chemotherapy-induced nausea and vomiting (CINV) [18]. So far, neither aprepitant nor casopitant have been formally tested for patients with radiation-induced nausea and vomiting in a randomized trial and can therefore not be recommended, although the pathogenesis of RINV is thought to be mediated in parts by substance P [47].

Duration of prophylaxis

The appropriate duration of 5-HT3 receptor antagonist prophylaxis for patients receiving fractionated radiotherapy is not clear. Although randomized trials have used 5-HT3 receptor antagonists either for extended periods [12, 36, 46] or just for the first five treatments, there have been no randomized trials that compare these approaches.

Rescue therapy

The role of antiemetics given on an as-needed basis has been investigated in three randomized trials [24, 30, 32].

In the latest study including 288 patients, the value of prophylactic versus rescue tropisetron was evaluated in fractionated radiotherapy of moderate or high emetogenic radiation fields [32]. Although no stratification in regard of radiation fields were conducted, it can be concluded that prophylactic tropisetron was superior over rescue tropisetron in regard to vomiting (risk 2.19 higher for rescue, p = 0.001) and nausea (risk 1.89 higher for rescue, p = 0.009). These results are in line with previous reported studies suggesting the benefit of 5-HT3 receptor antagonists even as a rescue medication for RINV.

In another study, 455 patients who developed emesis and/or moderate/severe nausea after receiving fractionated radiotherapy to sites located between the thorax and pelvis were randomized to receive one of the following treatments: (1) two placebo orally disintegrating tablets (odt); (2) one 8-mg OND odt and one placebo odt (OND 8 mg group) and (3) two 8 mg OND odt (OND 16 mg group) [24]. The study showed that OND was clinically superior to placebo independent of the OND dose prescribed. Both OND 8 and 16 mg doses increased treatment success over the 12 h after treatment compared with placebo (53% and 56%, respectively, compared with 41% for placebo). Statistically significant differences were observed between OND 8 mg and placebo (p = 0.026) and between OND 16 mg and placebo (p = 0.008). There was no significant difference between the two doses of OND. This trial showed once again that a 5-HT3 receptor antagonist, in this case ondansetron, given as a rescue treatment can control emesis and nausea effectively in radiotherapy patients who have already established symptoms [24].

The IGARR study by Maranzano et al. [28] conducted a double-blind randomized clinical trial in patients undergoing fractionated radiotherapy to the upper abdomen comparing prophylactic OND plus DEX versus placebo and, in case nausea and vomiting was observed, a cross-over rescue treatment. Of the 400 patients planned as the sample size to reach in three consecutive years, only 155 patients entered the trial, of which 153 patients were evaluable. Comparing OND plus DEX prophylaxis with placebo, vomiting was observed in 30% versus 40% and nausea in 57% versus 67% of patients, respectively. The 10% of major control in the group treated with OND plus DEX was not statistically significant. In regard to the rescue treatment, the placebo given to the group of patients who developed vomiting in spite of the prophylaxis with OND plus DEX was effective in 24% of cases, whereas when vomiting was not controlled with placebo given as a prophylactic drug, the rescue with OND plus DEX was effective in 64% of cases. Thus, when a rescue treatment was used, OND plus DEX was significantly more effective than placebo in controlling vomiting (64% versus 24%, p = 0.003). On the contrary, nausea was controlled in 56% of cases in both rescue treatment groups. In conclusion, this study showed that (a) in clinical practice, radiation oncologists generally underestimate the problem of nausea and vomiting in radiotherapy; and (b) antiemetic rescue treatment seems to be also effective [28].

In the observational study by Maranzano et al., it could be shown that antiemetic drugs were administered in 17% of the patients receiving radiotherapy, from which 12.4% were given prophylactically and 4.6% on a rescue basis [29].

Two open pilot studies evaluated the use of rescue treatment using a 5-HT3 receptor antagonist in patients failing to achieve relief with non-5-HT-3-receptor-antagonist-antiemetics [31, 38]. In the first study, four patients who had RINV after prophylaxis with prochlorperazine and metoclopramide received rescue treatment with OND. All patients achieved complete protection from vomiting [38]. In the second trial, 34 patients experiencing RINV during fractionated radiotherapy to the abdomen were treated with tropisetron, which controlled vomiting in 73% of cases [31]. The beneficial role of 5-HT3 receptor antagonists as rescue medication has been suggested in all these reports.

The role of rescue medication should be further explored in the low and minimal risk setting of RINV as the incidence of nausea and vomiting in the moderate and high emetogenic risk group is rather high, and a prophylactic use of antiemetics is mandatory in this setting.

Evaluating an antiemetic strategy in RINV using a risk factor model

In order to achieve an optimal treatment strategy to prevent nausea and/or vomiting, it could be useful to develop a risk-adjusted treatment for RINV. Therefore, the individual risk of the patient to develop nausea and/or vomiting should be taken into consideration as well as the emetogenicity of the radiotherapeutic regimen and any simultaneous administration of chemotherapy [7, 9].

Patient factors are known to influence the risk of emesis in cancer patients. For example, previous chemotherapy-induced emesis is a significant prognostic factor for developing RINV (IGARR [16]). Individual well-known risk profiles according to patient-related emetogenic risk factors are age, gender, alcohol consumption, previous experience of nausea and vomiting and anxiety (Fig. 1) [7, 9, 17, 27]. The emetogenic potential of radiotherapy is divided into high, moderate, low and minimal in a similar fashion to the emetogenicity of chemotherapy. Although not prospectively proven, an approach taking the known risk factors into considerations to tailor a risk-adjusted antiemetic treatment for RINV might be advantageous for patients. With this strategy, it might be possible to further improve the protection rate in patients receiving RINV (Fig. 2).

Fig. 1
figure 1

Individual risk factors according to patient-related emetogenic risk factors [6, 7, 9]

Full size image
Fig. 2
figure 2

Risk evaluation of emetogenic potential [6, 7, 9]

Full size image

New recommendations and conclusions

After the last MASCC Consensus Conference in 2004, new data on CINV and RINV suggested a need to update the existing guidelines. Therefore, MASCC convened an Expert Panel for a new International Antiemetic Consensus Conference which was held in Perugia, Italy, 20–22 June 2009. Data from the literature were evaluated, and relevant data with evidence of levels I and II were included. These provided the basis for the new proposed MASCC/ESMO guidelines discussed by the experts and described below.

According to the irradiated area (the most frequently studied risk factor); the guidelines are divided into four risk levels: high, moderate, low and minimal emetogenic risk of radiotherapy. The updated guidelines are shown in Table 3.

Table 3 Radiotherapy-induced emesis: radiation emetic risk levels and new MASCC and ESMO guidelines
Full size table

High risk

Total body irradiation (TBI) is associated with a high risk of RINV. In patients receiving TBI, a prophylaxis with a 5-HT3 receptor antagonist is recommended (MASCC level of confidence: high/level of consensus: high). The addition of dexamethasone to the 5-HT3 receptor antagonists has not been studied in this setting. However, if this approach adds efficacy, as occurs with chemotherapy, such regimen would be appropriate for patients submitted to TBI (MASCC level of confidence: low/level of consensus: high).

Moderate risk

Radiation of the upper abdomen, half body irradiation (HBI) and upper body irradiation (UBI) is associated with a moderate risk of RINV. In patients receiving radiotherapy with moderate emetogenic risk, a prophylaxis with a 5-HT3 receptor antagonist and optional in combination with a short course (day 1–5) of dexamethasone is recommended (MASCC level of confidence: high/level of consensus: high for the 5-HT3 receptor antagonist, when combined with dexamethasone: MASCC level of confidence: moderate/level of consensus: high).

Low risk

Radiation of the cranium (all), craniospinal radiation, H & N, lower thorax region and pelvis is associated with a low risk of RINV. In patients receiving radiotherapy with low emetogenic risk, a prophylaxis or a rescue therapy with a 5-HT3 receptor antagonist is recommended (for prophylaxis: MASCC level of confidence: moderate/level of consensus: high; for the 5-HT3 receptor antagonist, for rescue: MASCC level of confidence: low/level of consensus: high).

Minimal risk

Radiation of extremities and breast is associated with a minimal risk of RINV. In patients receiving radiotherapy with low emetogenic risk, a rescue with a dopamine receptor antagonist or prophylaxis with a 5-HT3 receptor antagonist is recommended (MASCC level of confidence: low/level of consensus: high).

Concomitant chemotherapy

In patients undergoing concomitant radiochemotherapy, the antiemetic prophylaxis should be according to the guidelines of chemotherapy-induced nausea and vomiting (CINV) choosing the emetogenic risk category of the used chemotherapeutic regimen. However, in case the risk category of radiotherapy is higher than the concomitant applied chemotherapy, the risk category of radiotherapy has to be chosen to tailor the antiemetic treatment (MASCC level of confidence: low/level of consensus: high).

The effect of RINV on patients’ quality of life should not be underestimated, especially as such effects may compromise or delay treatments. Therefore, patients at risk of RINV should always be offered the most effective antiemetic prophylaxis as suggested by this updated MASCC/ESMO 2009 guidelines.

Furthermore, there is an additional need to investigate the importance of the individual risk factors of patients, the incidence of delayed nausea and vomiting, and the potential role of NK-1 receptor antagonists, the 5-HT3 receptor antagonists palonosetron and transdermal granisetron patch as well as the duration of antiemetic treatment.

Note

MASCC evidence

Level of confidence (high–low) and grade of consensus (high–low) as used by the Multinational Association of Supportive Care are given in brackets.

ESMO evidence

Levels of evidence (I–V) and grades of recommendation (A–D) as used by the American Society of Clinical Oncology are in given in brackets.