In the United States, approximately 25 million surgical procedures are performed yearly (1). Postoperative nausea and vomiting (PONV) are common complications occurring up to 48 h after the administration of an anesthetic, with an incidence of 30%–70% (2). PONV results in more expensive health care costs secondary to unplanned admissions and expenses incurred from prophylaxis and treatment. Much research regarding PONV examines the incidence of nausea or vomiting in a given patient population, or between groups of similarly anesthetized patients given different prophylaxis regimens (3–5). Comparisons in these studies are often based on subjective nausea scores and/or actual incidence of retching (6–9).
The goal of the present study was to provide more definitive evidence that increases in plasma arginine vasopressin (AVP) concentration could be used to assess the presence of nausea produced by drugs dispensed postoperatively and whether frequently used antiemetic drugs have a demonstrable effect on plasma AVP concentration. Prior studies using a variety of emetogenic stimuli (e.g., motion, chemotherapy, and IV AVP) all indicate that AVP is associated with the genesis of nausea and vomiting. Increases and decreases in AVP correspond to the onset and resolution of nausea and vomiting, respectively. No studies, however, demonstrate whether the use of an antiemetic to attenuate nausea and vomiting correlates with the AVP concentration.
Experiments conformed to the National Institutes of Health Guide for the Care and Use of Experimental Animals (DHEW Publication No. [NIH] 85-23, Revised 1996, Office of Science and Health Reports, DRR/NIH, Bethesda, MD 20205) and were approved by the University of Pittsburgh's Institutional Animal Care and Use Committee. Experiments were performed in 14 male ferrets (weight range: 1.5–2.5 kg) obtained from Marshall Farms (North Rose, NY). Animals were housed singly in a light- and temperature-controlled room.
After a 5-day acclimation period in the animal housing facility and acclimatization to twice-daily handling, ferrets underwent implantation of a chronic indwelling cannula inserted into the left external jugular vein. This surgical procedure was performed aseptically in a dedicated operating suite on animals anesthetized using an IM injection of ketamine (25 mg/kg) and xylazine (2.5 mg/kg). Cannulae were made of silicone tubing (Esco Rubber; outside diameter 1.5 mm, inside diameter 0.5 mm) and were attached to an injection port that was sutured to the dorsal nuchal muscles and overlying skin. Animals were administered ketoprofen 15 mg/kg IM on postoperative days 0, 1, and 2. Catheter patency was maintained by twice-daily flushing with 1 mL of heparinized (1 U/mL) 0.9% NaCl solution. After the completion of blood sampling, animals were killed using an IV injection of sodium pentobarbital (125 mg/kg). Drug injection and sampling procedures were initiated on postoperative day 5 and were repeated on postoperative day 12, if the catheters remained patent and the ferrets were afebrile.
The effects of IV administered morphine sulfate with and without ondansetron pretreatment on plasma concentrations of AVP were determined in 10 animals. The cannulae in the other 4 animals did not remain patent long enough for blood sampling to occur.
Animals were gently restrained in one hand while a 1.0-mL baseline blood sample was withdrawn. This was followed immediately by an infusion of 1 mL of either plain 0.9% NaCl or 0.9% NaCl containing 0.1 mg/kg ondansetron (Glaxo Wellcome, Research Triangle Park, NJ). Twenty minutes later, another 1-mL blood sample was drawn in the same manner, followed immediately by an infusion of 0.1 mg/kg morphine sulfate (Elkins-Simm, Cherry Hill, NJ). Further 1-mL blood samples were withdrawn from each animal at 5, 10, 15, 30, 45, 60, and 90 min after infusion of the morphine sulfate. The blood volume removed was immediately replaced with 2 mL of 0.9% NaCl after each blood sample. Blood samples were drawn over 15–30 s and drug infusions and volume replacement were made over 30-45 s. Between blood samples, the animals were returned to their home cages and their behavior was constantly monitored. Provided the catheters remained patent, each animal had two sample sets of blood drawn: one with and one without ondansetron pretreatment. Blood sampling was separated by at least 7 days. Each study animal received both treatments. Alternate animals received ondansetron + morphine injections first followed by morphine alone or the opposite ordering of injections (i.e., morphine alone first followed by ondansetron + morphine).
Blood samples were collected into 2.0-mL microcentrifuge tubes (Fisher Scientific, Pittsburgh, PA), immediately immersed in an ice water slurry, and centrifuged within 60 min of collection (17,000g for 5 min). Plasma samples were removed and frozen at −80°C. Before assay, plasma samples were thawed on ice and extracted using C18 Sep-Pak Vac cartridges (1 mL, 50 mg; Waters, Milford, MA), as described previously (10). Each extract was frozen, dried using a Speed Vac (Savant Instruments), and then dissolved in 700 μL of buffer (50 mM NaPO4, 25 mM EDTA, 0.9% NaCl, 0.5% bovine serum albumin, 0.1% sodium azide). AVP concentrations were measured by radioimmunoassay in separate 200-μL aliquots of each plasma extract. The initial incubation, performed at 4°C for 48 h, contained 4500 cpm of 125I-labeled AVP (NEN-Dupont, Boston, MA) and a rabbit polyclonal antibody to AVP (Dr. J. Fernstrom, University of Pittsburgh, PA) in a volume of 400 μL. After incubation, antibodies were precipitated using a second antibody procedure: after overnight incubation with normal serum (1:600; Jackson Immunoresearch, West Grove, PA) and goat anti-rabbit immunoglobulin G (1:120; Linco, St. Charles, MO), tubes were centrifuged (3000g, 30 min), the supernatant fluid was aspirated, and the remaining pellets were counted in a γ counter (Packard-Auto Gamma Scintillation Spectrometer). Concentrations of AVP were calculated from standard curves generated with known amounts of synthetic peptide (Bachem, Torrance, CA) that were run through the same extraction procedure. All plasma samples were extracted at the same time and run together in the same assays for AVP. Assay sensitivity for AVP was 2.5 pg/mL. The observer was not blinded to treatment group.
Baseline plasma concentrations of AVP are variable in both humans and ferrets. Because of this variability, AVP concentrations were expressed as a ratio of the baseline value. For example, a value of 2 would indicate that the serum AVP concentration at that particular time was twice that of the animal's baseline AVP concentration. To determine whether administration of morphine only or morphine plus ondansetron pretreatment caused a statistically significant increase in plasma AVP as compared with baseline, the data were analyzed with a nonparametric one-way analysis of variance (ANOVA) test (Kruskal-Wallis test) combined with a Dunn's multiple comparison post hoc test. A Mann-Whitney test was used to determine whether baseline levels of AVP differed in the morphine or morphine plus ondansetron treatment groups. A two-way ANOVA procedure combined with Bonferroni posttests determined whether pretreatment with ondansetron altered the increase in AVP that occurred over time after administration of morphine. Statistical significance was set at P < 0.05. Pooled data are presented as means ± sd.
Fourteen animals were surgically implanted with IV catheters. The implanted catheters failed to function properly in four animals. They were killed before data collection. The 10 remaining animals received injections of either morphine alone (n = 10) or ondansetron and morphine (n = 7). Data from 1 animal that received both treatment regimens were excluded because the baseline AVP values were approximately 10 times larger than in the other animals, possibly because of an artifact induced during blood sampling or analysis. One animal developed a fever before first injection and was excluded from study. In two animals, there was a failure to withdraw blood between treatments. The final number of animals providing both sets of data was 6.
Figure 1 illustrates the changes in plasma AVP concentrations from baseline after administration of morphine in all 6 animals. The plasma AVP concentrations increased immediately after morphine administration and were significantly larger than baseline at 45, 60, and 90 min after the injection of morphine. Over the same time period, 5 of the 6 animals exhibited signs and symptoms of nausea including licking and retching. Figure 1. Comparison of vasopressin (AVP) level after morphine versus morphine plus ondansetron injection. Data from six animals were pooled. Ondansetron pretreatment resulted in significantly lower AVP levels at 60 and 90 min than injection of morphine alone. Results are expressed as means and standard deviations.
A week later, we tested whether administration of ondansetron 20 min before injection of morphine altered the increase in plasma AVP concentrations produced by morphine. At the time morphine was delivered in this subset of animals, plasma AVP levels were not significantly different (P = 0.38, Mann-Whitney test) from the previous study in which ondansetron pretreatment was not used. The baseline AVP values were 2.4 ± 0.5 pg/mL in the animals only given morphine, and 3.8 ± 2.4 pg/mL in the animals pretreated with ondansetron. Figure 1 also shows the effects of morphine injection on plasma AVP levels at different time points in the animals pretreated with ondansetron. A one-way ANOVA failed to demonstrate a significant increase in AVP levels above baseline (P = 0.46) when morphine was injected subsequent to ondansetron. In addition, a two-way ANOVA confirmed that ondansetron significantly blunted the increase in AVP levels induced by injection of morphine at 60 and 90 min (P < 0.05). Only 1 of the 6 animals injected with morphine subsequent to ondansetron exhibited any behavioral indicators of nausea, as opposed to 5 of 6 animals given morphine alone.
The major finding of this study is that the increase in plasma AVP concentrations associated with injection of morphine is blunted by pretreatment with ondansetron. These findings show that monitoring plasma AVP levels may, after further study, provide a means to predict the occurrence of vomiting (11). Thus, monitoring plasma hormone concentrations might have diagnostic value in determining which postsurgical patients are in need of treatment to avoid emesis. The ability to reliably and objectively measure nausea markers in a timely manner may ultimately reduce health care spending and improve patient satisfaction.
Ferrets are among the most reliable models for emetic research (12). A ferret model was used in the first studies regarding the effectiveness of 5-hydroxytryptamine (5-HT)3 antagonists in the late 1980s and early 1990s (13–15). In addition, prior work has shown that administration of large doses of one powerful nauseogenic drug, cholecystokinin, can produce an increase in AVP levels in ferrets (16). The ferret model was used to determine both the effectiveness of a novel 5-HT3 antagonist (ondansetron) and also some preliminary dosing. The doses of both morphine and ondansetron used in the Thompson et al. (12) study are supra-therapeutic. The goal of this study was to determine the effect of the drugs in a clinically relevant dose. We know that a dose of 0.1 mg/kg morphine would induce most ferrets to vomit, and, from this study, a dose of ondansetron 0.1 mg/kg would seem to prevent most ferrets from vomiting if given in a timely manner. Unfortunately, there are no good data to suggest exactly what time ondansetron should be given in relationship to the morphine dose.
Several studies have indicated a relationship between AVP levels and nausea and vomiting. Caras et al. (17) showed a correlation between the serum level of AVP and the intensity of nausea in a study in which AVP (either 0.15 U·kg−1·h−1 or 0.3 U·kg−1·h−1) was infused into 5 healthy patients. Symptoms of nausea were correlated with the infusion rate of AVP and increased significantly when larger doses of AVP were administered. Kim et al. (18) compared nausea elicited by either a motion stimulus or an IV AVP infusion. Both treatment groups experienced similar occurrences of nausea, and the motion stimulus group also showed a marked increase in plasma AVP levels. Another study by Barreca et al. (19) examined the levels of AVP in response to the administration of adjuvant chemotherapy. Seven of nine patients experienced nausea and/or emesis, and their AVP levels were sig-nificantly increased compared with the non-nauseated patients whose serum AVP levels did not increase. Friess et al. (20) observed a significant (14- to 26-fold) increase in baseline AVP levels in nauseated hypotensive patients as compared with non-nauseated patients.
Page et al. (21) examined the relative increase in serum AVP after the administration of syrup of ipecac. Their study revealed that there was a direct correlation between the nausea score and the peak incremental AVP response (P < 0.05). Interestingly, this study also demonstrated a decrease in AVP with the resolution of symptoms. Xu et al. (22) examined AVP concentrations in patients undergoing a motion-sickness stimulus, and similar to Page et al.'s (21) findings, serum AVP levels increased within 6 minutes of the beginning of the experiment and decreased as the nausea resolved. Patients experiencing nausea had significant increases in serum AVP compared with both their baseline levels, and the levels of the non-nauseated patients (P < 0.04). This led them to conclude that AVP concentrations correlate most closely with temporal onset and resolution of nausea. Edwards et al. (23) found that patients who vomited had increases in serum AVP from 4-129 Author: Please clarify numbers. fold. One patient was studied as both a vomiter and a nonvomiter to the chemotherapy stimulus. As a vomiter, this patient experienced a 16-fold increase in serum AVP and no increase as a nonvomiter.
A variety of serum hormone responses that may reflect nausea have been assayed in several species, including cholecystokinin and oxytocin (24,25). These substances are all readily measurable using the same plasma radioimmunoassay technique used in this study. The monitoring of multiple hormones, as opposed to a single agent, may provide more accurate predictions of whether nausea is present.
AVP concentration increases as a response to stress in both humans and animals. Although the blood volumes removed from our study animals were <2% of total blood volume of the animals, and the blood was replaced with saline in a 3:1 ratio, the possibility exists of a confounding increase in AVP levels secondary to induced hypovolemia. If this were the case, one would have expected a statistically significant increase in the AVP levels in the ondansetron pretreatment group of animals, which did not occur. Nonetheless, hypovolemia and other forms of stress could induce increases in AVP levels that could erroneously signal the presence of nausea. Combining assays of AVP with other markers for nausea could potentially overcome this problem.
In summary, this study demonstrates that serum AVP levels can serve as a marker for nausea and vomiting, including that induced by drugs frequently used during the administration of an anesthetic. Further studies are warranted to develop simpler assays for AVP and perhaps other hormones that reflect the presence of nausea, so that they can be put into common practice in the recovery room.
The authors acknowledge Alan Sved, PhD, Department of Neuroscience, University of Pittsburgh, Pittsburgh, PA for invaluable help with vasopressin assays.
1. Kapur P. The big “little problem” [editorial]. Anesth Analg
2. Camu F, Lauwers MH, Verbessem D. Incidence and aetiology of post-operative nausea and vomiting. Eur J Anaesthesiol
3. Watcha MF, Smith I. Cost-effectiveness analysis of antiemetic therapy for ambulatory surgery. J Clin Anesth
4. Fujii Y, Tanaka H, Somekawa Y. Granisetron, droperidol, and metoclopramide for the treatment of established postoperative nausea and vomiting in women undergoing gynecologic surgery. Am J Obstet Gynecol
5. Wittels B, Faure EA, Chavez R. et al. Effective analgesia after bilateral tubal ligation. Anesth Analg
6. Fujii Y, Toyooka H, Tanaka H. Prevention of postoperative nausea and vomiting in female patients during menstruation: comparison of droperidol, metoclopramide and granisetron. Br J Anaesth
7. Polati E, Verlato G, Finco G. et al. Ondansetron versus metoclopramide in the treatment of postoperative nausea and vomiting. Anesth Analg
8. Raphael JH, Norton AC. Antiemetic efficacy of prophylactic ondansetron in laparoscopic surgery: randomized, double-blind comparison with metoclopramide. Br J Anaesth
9. Alon E, Himmelseher S. Ondansetron in the treatment of postoperative vomiting: a randomized, double-blind comparison with droperidol and metoclopramide. Anesth Analg
10. Billig I, Yates BJ, Rinaman L. Plasma hormone levels and central cFos expression in ferrets after systemic administration of cholecystokinin. Am J Physiol Regul Integr Comp Physiol
11. Koch KL, Summy-Long J, Bingaman S. et al. Vasopressin and oxytocin responses to illusory self-motion and nausea in man. J Clin Endocrinol Metab
12. Thompson PI, Bingham S, Andrews PL. et al. Morphine-6-glucuronide: a metabolite of morphine with greater emetic potency than morphine in the ferret. Br J Pharmacol
13. D'Allens H, Aubert B, Bons J, Pappo M. Ondansetron: a specific 5-HT3 serotonin receptor inhibitor, a new anti-emetic in oncology. Bull du Cancer
14. Bermudez J, Boyle EA, Miner WD, Sanger GJ. The anti-emetic potential of the 5-hydroxytrypamine3 receptor antagonist BRL 43694. Br J Cancer
15. Kirchner V, Aapro M, Alberto P, et al. Early clinical trial of MDL 73.147 EF: a new 5-HT3-receptor antagonist for the prevention of chemotherapy-induced nausea and vomiting. Ann Oncol
16. Parrott RF, Forsling ML. Central and peripheral doses of cholecystokinin that inhibit feeding in pigs also stimulate vasopressin and cortisol release. Exp Physiol
17. Caras SD, Soykan I, Beverly V, et al. The effect of intravenous vasopressin on gastric myoelectrical activity in human subjects. Neurogastroenterol Motil
18. Kim MS, Chey WD, Owyang C, Hasler WL. Role of plasma vasopressin as a mediator of nausea and gastric slow wave dysrhythmias in motion sickness. Am J Physiol
1997;272(4 Pt 1):G853–62.
19. Barreca T, Corsini G, Cataldi A, et al. Effect of the 5-HT3 receptor antagonist ondansetron on plasma AVP secretion: a study in cancer patients. Biomed Pharmacother
20. Friess U, Rascher W, Ritz E, Gross P. Failure of arginine vasopressin and other pressor hormones to increase in severe recurrent dialysis hypotension. Nephrol Dial Transplant
21. Page SR, Peterson DB, Crosby SR, et al. The responses of arginine vasopressin and adrenocorticotrophin to nausea induced by ipecacuanha. Clin Endocrinol
22. Xu LH, Koch KL, Summy-Long J, et al. Hypothalamic and gastric myoelectrical responses during vection-induced nausea in healthy Chinese subjects. Am J Physiol
1993;265(4 Pt 1):E78–84.
23. Edwards CM, Carmichael J, Baylis PH, Harris AL. Arginine vasopressin: a mediator of chemotherapy induced emesis? Br J Cancer
24. Nussey SS, Hawthorn J, Page SR, et al. Responses of plasma oxytocin and arginine vasopressin to nausea induced by apomorphine and ipecacuanha. Clin Endocrinol
25. Ebenezer I, Thornton SN, Parrot RF. Anterior and posterior pituitary hormone release induced in sheep by cholecystokinin. Am J Physiol