Postoperative nausea and vomiting (PONV) continues to be a major problem although avoiding PONV is a high priority for patients and physicians. Apart from patient discomfort, PONV is associated with adverse effects, such as delayed recovery and prolonged hospital stay. Although rare, postoperative morbidities including wound dehiscence, pulmonary aspiration, bleeding, and dehydration that can occur if vomiting is prolonged [1]. Established patient-specific risk factors for PONV include female gender, nonsmoking, and a history of motion sickness or PONV, whereas nonspecific factors are postoperative opioids use and type of surgery such as laparoscopy [2-4].

Ondansetron was the first commercially available 5-hydroxytryptamine (5-HT₃) receptor antagonist, thereafter granisetron, dolasetron, tropisetron, ramosetron, and palonosetron were introduced. Many studies have confirmed that this class of antiemetics exhibited better prophylactic efficacies compared with the older traditional drugs including droperidol, perphenazine, or metoclopramide [5-7]. A number of studies have been conducted to compare the prophylactic antiemetic efficacies among 5-HT₃ receptor antagonists [8-12]. To date, however, there are few clinical studies comparing the prophylactic efficacies of ondansetron, ramosetron, and palonosetron in high-risk patients with PONV.

The aim of this prospective, randomized, double-blinded trial was to compare the prophylactic antiemetic efficacies of ondansetron, ramosetron, and palonosetron in high-risk patients with fentanyl-based PCA after laparoscopic surgery. Our hypothesis was that three 5HT₃ receptor antagonists show different antiemetic efficacies when used prophylactically.

Our study was conducted from July 2010 to June 2011 and 109 patients were enrolled. The inclusion criteria were as follows: 1) American Society of Anesthesiologists physical status I-II, 2) female patient aged 20-65 years, 3) nonsmoker, 4) history of motion sickness or previous PONV, 5) elective laparoscopic surgery, and 6) use of postoperative intravenous patient-controlled analgesia (IV-PCA) using fentanyl. The exclusion criteria were: impairment of bowel motility, diabetes, pregnancy or lactation, administration of an antiemetic medication or steroids within 24 h before surgery, the presence of a cardiovascular or respiratory disease, obesity (body mass index > 35 kg/m²), or renal or hepatic dysfunction. The plan of the study was reviewed and approved by the Institutional Review Board of Yonsei University Health Systems) and written informed consent was obtained from all patients.

No patient withdrew from the study. Neither patient characteristics nor intraoperative data differed among the 3 groups (Table 1).

The number of nausea-free patients was greater in the palonosetron group than in the other 2 groups during the 48 h study period (Table 2). The number of nausea-free patients was also greater in the ramosetron group than in the ondansetron group overall. During the first hour after surgery, the overall number of patients who experienced postoperative nausea was low and not significantly different among the groups. The rescue antiemetic was used less frequently in the palonosetron group than in the other 2 groups. Patients who were given the rescue antiemetic were censored and Kaplan-Meier analysis of the interval to first rescue antiemetic showed a significant intergroup difference (Fig. 1) between the palonosetron and ondansetron (P < 0.001), palonosetron and ramosetron (P < 0.001), and ondansetron and ramosetron (P = 0.042) groups. Thus, the order of prophylactic efficacy in prolonging the time to use a rescue emetic was palonosetron, ramosetron, and ondansetron.

Six patients (17.1%) in the ondansetron group, 5 (14.3%) in the ramosetron group, and none (0%) in the palonosetron group (P = 0.041) requested disconnection of the IV-PCA pumps because of intractable nausea and/or vomiting after surgery mostly within 12 h.

Pain scores and the total amounts of fentanyl administered via IV-PCA were similar in the three groups during 48 h after surgery (Fig. 2). The 3 groups were also similar in terms of the number of patients who experienced adverse events postoperatively (Table 3).

This prospective, randomized, double-blinded trial compared the prophylactic antiemetic efficacies of ondansetron, ramosetron, and palonosetron in high-risk patients. Palonosetron was superior to ondansetron or ramosetron when used prophylactically to reduce the incidence and severity of PONV and to delay the use of rescue antiemetics.

Consistent with previous findings [9], we observed very high incidences of PONV (80%) despite the prophylactic use of ondansetron. This is mainly due to the patient's factor. All study patients were at high-risk because they had at least 4 predictors of PONV: female gender, nonsmoker, use of postoperative opioids, and history of motion sickness or previous PONV. The incidence of PONV increases exponentially from 10% when no risk factor is present to 79% when 4 risk factors are all present [4,14]. Laparoscopic surgery is also highly susceptible to PONV because abdominal gas insufflation may stretch mechanoreceptors of the intestine and consequently activate 5-HT₃ receptors via serotonin release [16]. Another explanation for the frequent PONV in our results is the use of fentanyl-based IV-PCA, especially the use of basal infusion with relatively low demand capacity of fentanyl in our settings. Adding a basal infusion to IV-PCA could be more convenient in surgery and more effective in reducing resting pain than bolus demand only, but it may be associated with a greater risk of PONV. In addition, other reasons for unsatisfactory antiemetic effects of ondansetron given prophylactically are the timing of administration [17], the dose of ondansetrone [18], and CYP2D6 alleles polymorphisms [19].

Although ondansetron has been used to prevent PONV [20,21], there are conflicting results about its effect on preventing IV-PCA related PONV, which were not clinically satisfactory [22,23]. In our results, there was no difference among the groups in terms of the number of patients who experienced nausea and emetic episodes in the first hour after surgery. However, after that period, ondansetron did not reduce fentanyl-based IV-PCA related nausea and emetic episodes. Ramosetron, a newer 5-HT₃ antagonist, has been reported to be an effective antiemetic in patients undergoing various types of surgery. The elimination half-life of ramosetron (5.8 ± 1.2 h) is longer than that of ondansetron (3.8 ± 1.0 h) [24]. Two recent, randomized studies compared the prophylactic antiemetic efficacies of ondansetron and ramosetron in patients receiving opioid-based PCA [9,10] The former of the 2 studies found that the incidence of nausea was similar in the 2 groups, but that ramosetron decreased the severity of nausea and the incidence of vomiting, resulting in a reduced need for antiemetic rescue treatment [9]. The latter of the 2 studies showed that the rate of PONV and the use of rescue antiemetics were lower in the patients who received ramosetron than in patients who received ondansetron [10]. The present study showed that the ramosetron group had more patients who presented nausea-free postoperatively compared with the ondansetron group. During first 1-6 h after surgery, fewer patients who had emetic episodes were also in the ramosetron group than in the ondansetron group, despite the similar severity of nausea. In addition, Kaplan-Meier analysis revealed that ramosetron significantly prolonged the time to use a rescue emetic compared with ondansetron, which is attributable to failure of prophylaxis.

Palonosetron, the latest 5-HT₃ receptor antagonist introduced in 2003, has been proven to be effective when used to prevent emesis associated with chemotherapy [25]. The minimum effective dose of palonosetron was 75 µg and the drug has been approved by the United States Food and Drug Administration (FDA) for PONV prophylaxis [26,27]. We found that palonosetron exerted a greater prophylactic effect than did either ondansetron or ramosetron. Kaplan-Meier analysis also showed that the efficacy of palonosetron was superior to that of the other 2 drugs. The greater efficacy of palonosetron may be attributable to the fact that the binding affinity to the 5-HT₃ receptor is 30 times higher than that of either ondansetron or ramosetron and/or to the extended half-life of palonosetron (approximately 40 h, thus 4-10-fold longer than that of the older antagonists) [28]. However, these properties do not entirely explain the higher efficacy of palonosetron relative to ondansetron or ramosetron. If efficacy was attributable to potency alone, ondansetron or ramosetron could be administered at higher doses. If half-life was all-important, other drugs with shorter half-lives could be injected more often. However, administration of ondansetron for more than 24 h after chemotherapy did not prolong the emesis-free interval afforded by the use of palonosetron [29], suggesting that the longer half-life of palonosetron does not explain the greater antiemetic efficacy of the drug when given within 24 h after surgery. Palonosetron may be unique in terms of allosteric interaction with and binding cooperativity to 5-HT₃ receptors [30]. Moreover, binding of palonosetron to the receptors and thereby inhibiting calcium influx is not easily reversible, suggesting that palonosetron uniquely triggers 5-HT₃ receptor internalization and induces prolonged inhibition of receptor function [30].

Our study has several limitations. First, we did not include a placebo control group to evaluate the baseline incidence of PONV as we considered it unethical to withhold prophylactic antiemetic drugs in patients at high risk for PONV. Second, patient satisfaction was not an end-point. Third, combination therapy has been shown to be more effective to treat or prevent PONV, especially in high-risk patients. For example, corticosteroids such as dexamethasone are frequently combined with agents of other drug classes, such as metoclopramide or 5-HT₃ antagonists, resulting in significant improvements in response. Future research should assess the use of drug combinations to prevent PONV in high-risk patients; the work should also evaluate drug cost-effectiveness. It is also necessary to assess the efficacy of antiemetic drugs in the treatment of established PONV.

In conclusion, a single intravenous injection of 75 µg of palonosetron was more effective in preventing PONV and reduced the need to use a rescue antiemetic compared to an injection of 4 mg of ondansetron or 0.3 mg of ramosetron in high-risk patients who were scheduled for laparoscopic surgery with fentanyl-based postoperative pain management.