Substance P (Neurokinin-1) Antagonist Prevents Postoperative Vomiting after Abdominal Hysterectomy Procedures

A total of 243 healthy, consenting American Society of Anesthesiologists physical status I or II, nonpregnant women presenting for total abdominal hysterectomy procedures were successfully enrolled in these sequential, multi-institutional review board–approved, randomized, double-blind studies. The dose-ranging studies (n = 86) were designed to compare CP-122,721 100 or 200 mg orally with placebo. In the interaction study (n = 157), the incidence of postoperative emesis was evaluated in women treated with CP-122,721 200 mg orally, ondansetron 4 mg intravenously, or a combination of both. In the dose-ranging studies, patients were enrolled at three sites, whereas five sites were used for the interaction study.

Patients were excluded from participating in these studies if they had received any antiemetic medications within 24 h before induction of anesthesia, had received an investigational drug within the past 30 days, had vomited or retched within the preceding 24-h period, or were more than 60% above their ideal body weight. All patients provided a detailed medical history, including history of previous PONV, motion sickness, or dizziness, alcohol or drug consumption, as well as the date of their last menstrual period.

In the dose-ranging studies, either (1) a placebo or CP-122,721 100 mg or (2) a placebo or CP-122,721 200 mg was administered orally 60–90 min before induction of general anesthesia according to a randomized, double-blind protocol with two separate placebo groups. All study medications were prepared by the manufacturer (Pfizer, Inc., Groton, CT) in identical capsules. The general endotracheal anesthetic technique was standardized and included thiopental 3–5 mg/kg, fentanyl 2–4 μg/kg, isoflurane 0.6–1.2%, morphine 0.05–0.15 mg/kg, muscle relaxants and reversal drugs, as well as nitrous oxide. Morphine, 1–2 mg intravenously, was also administered for postoperative pain control using a patient-controlled analgesia (PCA) delivery system. The PCA morphine was initiated in the postanesthesia care unit when the patient first requested pain medication.

In the interaction study, patients were randomly assigned to one of three treatment groups using a computer-generated random number table. Each group received either CP-122,721 200 mg or placebo orally 60–90 min before induction of anesthesia, and either ondansetron 4 mg or saline 2 ml intravenously 15–30 min before the end of the surgical procedure under a standardized general anesthetic technique. Group 1 (ondansetron) received oral placebo and intravenous ondansetron, group 2 (CP-122,721) received oral CP-122,721 and intravenous saline, and group 3 (combination) received oral CP-122,721 and intravenous ondansetron. The oral study drug was supplied by Pfizer Inc. in blinded containers, whereas the parenteral (intravenous) solution was prepared by a local pharmacist who was not otherwise involved in the conduct of the study. The patients, anesthesiologists, and observers were all blinded to the study medications being administered.

The length of surgery (from skin incision to placement of the last suture) and anesthesia (from induction to discontinuation of nitrous oxide), as well as recovery times (from the end of surgery to eye opening and orientation to person and place, and discharge from the postanesthesia care unit) were also recorded. The degree of nausea and pain was assessed on arrival to the postanesthesia care unit and at 0.5-, 1-, 1.5-, 2-, 4-, 8-, 12-, and 24-h intervals after surgery using 100-mm linear visual analog scales (VAS; 0 = none, 100 = most severe). In the interaction study, patient satisfaction with the management of their emetic symptoms was also evaluated at 72 h postoperatively. The assessment involved asking the patients if they were “highly satisfied” with the management of their PONV symptoms. In addition, the occurrence of all clinically significant postoperative side effects (including changes in electrolytes and liver function tests), emetic episodes, and the need for rescue medications were recorded during the entire 72-h study period.

An emetic episode was defined as a single occurrence of vomiting or retching, and repeat episodes had to be separated by at least 2 min. In the postanesthesia care unit, a rescue medication (droperidol) was administered if the patient complained of persistent nausea (with a VAS nausea score > 40 for at least 10 min) or experienced at least two emetic episodes. If the patient required antiemetic (or antinausea) drug therapy during the 72-h study period, they were considered a treatment failure. All rescue antiemetic (or antinausea) medications, PCA morphine requirements, and other postoperative medications were documented. The sample size calculation for the initial placebo-controlled, dose-ranging study was based on a power analysis assuming a > 30% overall occurrence of postoperative emesis and the hypothesis that CP-122,721 would reduce the incidence by 50%. For the second interaction study, a minimum sample size of 40 patients for each group was determined by an a priori  power analysis based on the assumptions that (1) the incidence of PONV in this patient population would be ≥ 30 %, 10,11(2) a 10% reduction in PONV would be clinically relevant, and (3) α= 0.05 and β= 0.2.

Categoric data were analyzed using the chi-square test, and continuous data were analyzed with one way analysis of variance. For normally distributed data, the between-group comparisons were performed by analysis of covariance with baseline scores designated as the covariate. Non-normally distributed data were analyzed using a Kruskal-Wallis test followed by Kruskal-Wallis Z test for multiple comparisons. Time to first emetic event and to rescue medications were analyzed using log-rank test statistics. The curves of the time to when 25% of the patients in each group were judged to have failed the prophylactic antiemetic therapy (i.e. , had their first episode of emesis–retching or required rescue antiemetic therapy for nausea) were determined by the Kaplan-Meier method. Summary statistics included mean values ± SD, median values and ranges, and percentages and numbers. A P  value < 0.05 was considered statistically significant. All statistical tests were performed using the Number Cruncher Statistical Systems version 6.0 program (NCSS Corp., Kaysville, UT).

Of the 277 patients enrolled in the two studies, 34 were withdrawn because of protocol violations (e.g. , inadvertent administration of a drug with antiemetic properties during surgery), failure to receive the study medication at the appropriate time, or cancellation of the surgical procedure. In both the dose-ranging (n = 86) and interaction (n = 157) studies, the treatment groups were comparable with respect to age, weight, height, history of PONV and motion sickness, day of menstrual cycle, duration of surgery, and PCA morphine usage during the first 72 h postoperatively (tables 1 and 2).

In the dose-ranging studies, both CP-122,721 100- and 200-mg oral doses were found to decrease emetic symptoms compared with placebo treatments (table 1). In addition, the 200-mg dose of CP-122,721 delayed the onset of emesis compared with the placebo treatment (fig. 1). Although the maximum nausea VAS scores were lower in both CP-122,721 groups compared with placebo treatments, these differences did not achieve statistical significance. In the early postoperative period, pain VAS scores were similar in patients treated with CP-122,721 100 or 200 mg orally compared with the placebo treatments. There was no significant difference between the PCA morphine requirement in the CP-122,721 and placebo groups during the initial 24-h postoperative period.

The VAS nausea scores did not differ among the ondansetron, CP-122,721, and the combination groups (table 3). Although the percentage of patients complaining of nausea (VAS score > 40) and the time to and requirements for rescue antiemetic drugs were similar in all three groups (fig. 2), the incidence of the emetic episodes was significantly lower after CP-122,721 200 mg alone or in combination with ondansetron 4 mg compared with ondansetron 4 mg alone (table 4). In addition, the Kaplan-Meier plot shows that the median time interval for 75% of the patients to remain completely free from any emetic episodes was significantly longer after the combination of CP-122,721 200 mg and ondansetron 4 mg than with either drug alone (fig. 3). Analogous to the findings in the placebo-controlled study, CP-122,721 was not associated with apparent analgesic- or opioid-sparing activity compared with ondansetron (table 4).

Compared with the placebo treatments, the only clinically significant adverse event attributed to CP-122,721 during the 72-h follow-up period was an increased incidence of headaches. Of 41 patients in the two CP-122,721 groups in the dose-ranging studies, 9 patients (22%) complained of postoperative headaches compared with only one patient (2%) in the placebo groups (P < 0.05). All of the headaches were either of mild (60%) or moderate (40%) severity. In the interaction study, there was a similar incidence of headaches in the ondansetron, CP-122,721, and combination groups: 1 (2%), 5 (10%), and 3 (6%), respectively. There were no differences in other side effects among the three groups in the second study. Finally, patient satisfaction with the management of their PONV symptoms was similar with CP-122,721 and ondansetron (table 3).

Gynecologic surgical procedures are associated with a high incidence of PONV. 11,12The incidence of PONV after lower abdominal gynecologic procedures without prophylactic antiemetic treatment varies from 40% to 80%. 1Ondansetron has been shown to decrease the incidence of PONV after gynecologic surgery, 12,13and recent studies have suggested that it is more effective when administered near the end of the surgical procedure. 4,10Therefore, in our second study, ondansetron was administered 15–30 min before the end of the operation in an effort to optimize its antiemetic efficacy in this high-risk gynecologic surgery population. Although the meta-analysis by Tramer et al.  6suggested that an 8-mg dose of ondansetron provided more effective prophylaxis against PONV than a 4-mg dose, this finding has been questioned because of the many inherent problems with meta-analysis, 14including a lack of standardization of the timing of drug administration, as well as differences in the anesthetic techniques and types of surgical procedures. 15A more recent dose-ranging study involving the use of ondansetron for PONV prophylaxis found that 4 mg administered intravenously was equivalent to 8 mg administered intravenously. 16Despite optimization of the antiemetic dosage regimen, up to 20% of patients undergoing high-risk gynecologic procedures will continue to experience emesis with ondansetron, consistent with our experience in the interaction study.

The vomiting reflex can be triggered by activation of both mechanoreceptors in the gastrointestinal tract as well as via  central mechanism during the perioperative period. A drug that can block this reflex arc should be useful in the prevention of nausea and vomiting. 1,2Substance P, an endogenous ligand for the NK-1 receptor, evokes a wide variety of biological responses, including stimulation of gastrointestinal smooth muscle activity, exocrine gland secretion, afferent sensory responses to gastric distention, and other visceral afferent stimuli. 17,18It is present in both vagal and sympathetic afferents and may potentiate “wind-up,” as well as the activation of reflexes mediated, in part, by other neurotransmitter systems. Animal studies showed that selective NK-1 antagonists have a broad-spectrum antiemetic effect that is dependent on their ability to penetrate the central nervous system. 7,8,17,18A preliminary study involving another NK-1 receptor antagonist (GR 205171) used postoperatively found that this investigational compound also provided better control of PONV than placebo in patients undergoing major gynecologic surgery procedures. 19 

Because NK-1 antagonists do not have activity at the 5-HT³receptors, 7–9we hypothesized that the combination of CP-122,721 and ondansetron would exert a more profound antiemetic effect because of their antagonist activity at different central nervous system receptor sites. The current study suggests that the antiemetic activity of this NK-1 antagonist compares favorably with ondansetron. Furthermore, the prolongation of the emesis-free period in the group receiving the combination therapy would suggest that blockade of both NK-1 and 5-HT³receptors may result in enhanced antiemetic activity. Further investigations of this selective NK-1 antagonist alone and in combination with other commonly used antiemetic drugs are necessary to determine the optimal use of this novel compound in clinical practice.

Tramer et al.  6reviewed 53 clinical trials that investigated the efficacy and safety of ondansetron for preventing PONV. These investigators concluded that patients receiving ondansetron for antiemetic prophylaxis were at risk for postoperative headaches. Although CP-122,721 was well tolerated and apparently safe when administered either alone or in combination with ondansetron, it was also associated with an increased incidence of postoperative headaches (22%vs.  only 2% in the placebo groups).

In a cost-effectiveness analysis of ondansetron that examined both the direct and indirect costs, Watcha and Smith 20concluded that routine prophylaxis with ondansetron was cost-effective if the expected incidence of PONV exceeded 30%. This estimate was based on the suboptimal efficacy obtained with existing dosing and timing regimens, and using the drug alone rather than in combination with other antiemetic agents. As pointed out by Fisher, 15a major problem with many of the early PONV studies involving ondansetron (and other newer 5-HT³antagonist drugs) is that they focused exclusively on so-called surrogate end points. Of interest, more recent publications involving the use of ondansetron as a prophylactic antiemetic have provided data on clinically relevant outcome measures (e.g. , patient satisfaction, time needed to resume normal activities, and willingness to pay). 4,11The failure to obtain recovery and outcome data represents serious deficiencies in the current study designs. The availability of these data in follow-up studies with CP-122,721 will make it possible to perform more meaningful comparative assessments of this novel antiemetic drug.

In conclusion, preoperative administration of the orally active NK-1 antagonist CP-122,721 (200 mg orally) was found to be similar to ondansetron (4 mg intravenously) in decreasing emetic symptoms after abdominal hysterectomy procedures. Patient satisfaction with the control of their PONV symptoms after CP-122,721 prophylaxis was also similar to ondansetron. The combination of CP-122,721 and ondansetron significantly prolonged the time to administration of the first rescue antiemetic drug compared with either drug alone, and almost completely prevented the occurrence of emesis. These preliminary data suggest that the combination of an NK-1 and 5-HT³receptor antagonist may be useful for antiemetic prophylaxis of surgical patients at high risk of developing PONV. NK-1 antagonists may represent a useful new class of antiemetic drugs.

The authors thank our colleagues for their cooperation with the anesthetic protocol used for these clinical investigations.

TABLE