New pharmacologic findings for the treatment of PONV and PDNV

A better understanding of pharmacologic data can help clinicians make more educated decisions, provide better patient care, and eventually lead to improved outcomes for patients with a risk of developing postoperative nausea and vomiting (PONV) or postdischarge nausea and vomiting (PDNV). There are several important negative consequences of PONV, including delayed ambulation, increased hospital stays,¹ dehydration, and interference with oral medications and nutrition.² Although severe negative outcomes associated with nausea and vomiting are rare (such as wound dehiscence and esophageal tears), further reduction of the incidence of nausea and vomiting lowers the likelihood of any potential medical complications occurring (Table 1).³ Furthermore, unplanned hospital admissions detract from achieving the goal of same-day discharges⁴ and result in inconvenience to patients and their families, lost wages and missed time at work, as well as increased costs to hospitals and possible lost revenue from the inability to accept new surgical patients if the inpatient surgical unit capacity is limited. Carroll and colleagues reported data on 211 patients from six hospital-based surgical centers who underwent ambulatory surgeries, including 34 patients who subsequently experienced PONV, demonstrating the potential financial impact of suboptimally managed PONV.¹ Their data indicated that an average of $415 per patient (1994 dollars) was incurred due to lost revenue and increased operating costs from a single episode of PONV, even without an unplanned admission.

There have been several milestones in the development of antiemetic treatments for cancer patients receiving emetogenic chemotherapy and data that have helped guide drug development for PONV. Members of the phenothiazine drug class were the first agents reported to have an antiemetic effect in a randomized controlled study. In 1963, Moertel and colleagues demonstrated that the monotherapies prochlorperazine or thiopropazate (an agent no longer available in the U. S.) were superior to placebo for reducing vomiting in colorectal cancer patients receiving single agent 5-fluorouracil chemotherapy.⁵ The introduction of cisplatin in the late 1970s stimulated renewed interest in antiemetic research due to the unprecedented nausea and vomiting that accompanied the extraordinary anti-tumor efficacy of this agent. In 1981, Gralla and colleagues described improved antiemetic efficacy with high-dose metoclopramide regimens as prophylaxis for highly emetogenic chemotherapies, thus suggesting a potential role in preventing PONV/ PDNV.⁶ Notably, during that same decade, combination therapy began to be regarded as an important way to enhance antiemetic efficacy; in particular, corticosteroids were commonly added to various agents from other drug classes, such as high-dose metoclopramide or a serotonin antagonist, resulting in significantly improved antiemetic responses.^7,8 During the mid-1980s, clinical development of 5-HT₃ antagonists led to the approval in 1991 of ondansetron, the first commercially available agent in this drug class for use in the U. S. In turn, this led to a great improvement of the care of chemotherapy patients due to the more potent antiemetic effect of ondansetron in the treatment of chemotherapy-induced nausea and vomiting (CINV).⁷ More recently, neurokinin (NK-1) antagonists have been added to the armamentarium and have provoked considerable interest. Following their clinical development during the 1990s, the first and, currently, only FDA-approved NK-1 receptor antagonist (aprepitant) was approved for prevention of chemotherapy-induced emesis in 2003 and in July 2006 by the FDA for the prevention of PONV.⁹

Decades of accumulated basic research have identified a plethora of mechanisms involved in regulating and processing the emetic reflex, which are also potential points of action for pharmacotherapies to prevent and treat nausea and vomiting.⁷ A summary of the known neurotransmitter-receptor systems involved in producing the emetic re-flex is described in Figure 1. Each of these pathways provides opportunities to reduce nausea and vomiting as they function independently of each other. Therefore, when therapies from multiple drug classes are combined (thereby targeting multiple receptor systems), an increase in antiemetic efficacy is generally observed, illustrating an additive protective effect on reducing nausea and vomiting (Table 2).¹⁰ However, interfering with these multiple endogenous pathways can also cause additional adverse effects, as most of these systems do not function solely to regulate the emetic response, rather they are also associated with other physiological functions. For example, dopamine is involved in regulating motor control as well as the emetic response, and drugs which block dopaminergic transmission (e.g., metoclopramide and prochlorperazine) can cause extrapyramidal reactions (e.g., dystonic reactions), anxiety, or depression.⁷

Overview of antiemetic agents currently in use

Antagonists of dopamine D₂ receptors

Droperidol and metoclopramide are prime examples of dopamine antagonists. Both have been demonstrated to be effective antiemetic agents, especially when used in combination with serotonin antagonists or dexamethasone, respectively.^11,12 Still, usage of metoclopramide, either as monotherapy or in combination, is not supported by the 2007 Society for Ambulatory Anesthesia (SAMBA) guidelines on PONV.¹³

Some antiemetic agents do not target a single receptor; therefore, they can affect more than one receptor system. For example, metoclopramide inhibits the dopaminergic system, but also has a fairly substantial effect on serotonin receptors as well.¹² Similarly, droperidol affects other systems beyond its antagonism of dopamine in the chemoreceptor trigger zone and area postrema which mediates its antiemetic effect.¹⁴ It is the drug’s effect on dopamine systems elsewhere in the brain that leads to sedation, dyskinesia, restlessness, and dysphoria. Droperidol’s impact on postsynaptic α1-adrenoceptors may cause hypotension, and its putative action on the pore-forming α-subunit of potassium ion channels may be the cause of its most serious and well-publicized adverse effect of QTc prolongation.^15,16 Despite the safe use of droperidol for treating CINV and PONV for more than 30 years,¹⁷ in 2001 the FDA issued a black-box warning on droperidol based on adverse effects observed during a psychiatric study of a 25 mg intravenous (i.v.) dose, as well as 65 individual case reports.¹⁸ Importantly, the black-box warning is based on the recommended and approved dosages of the drug listed on the package insert,¹⁹ which are actually higher than the amounts typically used in clinical practice to treat nausea and vomiting. In fact, of the information which influenced the FDA decision, only two of the case reports listed dosages similar to those administered in the U.S., while the European studies used doses 50 to 100 times higher than the antiemetic doses of droperidol generally used in clinical practice; 0.625 mg to 1.25 mg i.v.^17,18 These doses, when used for PONV, have been demonstrated to provide superior efficacy to placebo and equal efficacy to ondansetron, as indicated by numbers of patients with a complete response (CR) to antiemetic therapy. These efficacy findings were not associated with any significant differences in adverse effect profiles.²⁰

Histamine Receptor Blockers

Histamine receptor blockers are a major class of antiemetic agents that block acetylcholine in the vestibular apparatus and the nucleus tractus solitarius (a region of the brain stem involved in coordinating the emetic reflex). From this drug family, promethazine is most commonly administered for treatment of PONV. The efficacy of the drug for PONV has been demonstrated by randomized controlled trials,²¹ and its use is supported by current guidelines.¹³ Promethazine targets muscarinic receptors and blocks dopamine receptors as well. The latter likely cause many of the adverse effects of promethazine, which, similar to droperidol, include sedation, restlessness, and extrapyramidal symptoms. Still, the larger problem with promethazine use is the local tissue necrosis reported after intraarterial and sometimes i.v. injection.²² The current recommendation is that promethazine should be significantly diluted and administered as an i.v. drip, although this method is eminently impractical for anesthesiologists.

Cholinergic receptor antagonists

The plant extract scopolamine is a classic member of the class of drugs that antagonize muscarinic cholinergic receptors located in the cerebral cortex and pons.¹⁰ Trans-dermal scopolamine is also recommended by the SAMBA guidelines,¹³ and the patch is an effective antiemetic as demonstrated in a trial of 48 gynecologic laparoscopy patients followed for 24 hours after surgery.²³ However, use of the drug can be limited by sedation and other anticholinergic adverse effects, such as somnolence, dry mouth, and, less commonly, dizziness.¹⁰

5-HT₃ receptor antagonists

The mainstay of current antiemetic treatment is the 5-HT₃ subclass of serotonin antagonists, because they are highly effective, highly selective for their target, and extremely well tolerated. Because the 5-HT₃ subtype of serotonin receptors does not play as significant a role in other brain functions that typically lead to adverse effects when disrupted, targeting these receptors does not commonly induce central adverse effects to the same extent as other drug classes. For example, sedation and extrapyramidal symptoms are not typically observed, although headache is one of the more commonly reported adverse effects.²⁴ Agents in this class commercially available in the U.S. include ondansetron, granisetron, dolasetron, and palonosetron.

Other agents

Other drug classes that have a role in preventing nausea and vomiting postoperatively or following chemotherapy include NK-1 receptor antagonists. An oral formulation of aprepitant was recently approved for the prevention of PONV. Recent clinical data include two double-blind trials with a total of 1599 participants who were undergoing major operations under general anesthesia (primarily gynecological surgery) demonstrating better efficacy with aprepitant compared to ondansetron.²⁵

Corticosteroids, such as dexamethasone, are first-line antiemetic agents in the U.S. Although their precise mechanism of preventing PONV is unknown, their effects are thought to be mediated by the drug class’s antiinflammatory or membrane-stabilizing activities peripherally or centrally.¹⁰

Cannabinoids have been demonstrated to be effective and hold promise for a role in treating CINV; however, the literature does not support their use for PONV,²⁶ nor do the 2007 SAMBA guidelines recommend them.¹³

Palonosetron versus other 5-HT₃ receptor antagonists

As previously mentioned, administering multiple drugs from a single drug class has not been found to be beneficial for preventing or treating PONV. Once an antiemetic agent has been given in a sufficient dose to block a specific receptor system, it has not been useful to give another compound from the same drug class. Yet, new data on palonosetron, a recently developed 5-HT₃ receptor antagonist, have begun to challenge this theory.

Palonosetron, a second generation 5-HT₃ receptor antagonist,²⁷ is approved by the FDA for the prevention of acute and delayed CINV at a dose of 0.25 mg i.v. (administered prior to chemotherapy) and as an injection for the prevention of PONV for up to 24 hours following surgery at a dose of 0.75 mg i.v. (administered over 10 seconds immediately before the induction of anesthesia). A large randomized, double-blind trial published in 2006 compared palonosetron with ondansetron in the prevention of CINV. The authors found that CR rates were slightly higher with palonosetron than ondansetron during the delayed (24–120 hours) and overall (0–120 hours) periods following administration of highly emetogenic chemotherapy (cisplatin > 60 mg/m²).²⁸ In addition, within the 447 patients pretreated with palonosetron 0.25 mg who also received dexamethasone, significantly higher CR rates were observed than those receiving ondansetron plus dexamethasone, during both the delayed (42.0% versus 28.6%) and overall (40.7% versus 25.2%) periods after chemotherapy.²⁸ The adverse effect profile was comparable to other drug class members; headache and constipation were the most commonly reported adverse events for both antiemetic agents.

Potentially, some of the observed differences between palonosetron and the other 5-HT₃ receptor inhibitors could be due to the structural characteristics of this molecule, which affect its pharmacokinetic and pharmacologic profile. A major structural difference between palonosetron and the other agents could be expected to impact antagonist binding to the 5-HT₃ receptor. In fact, palonosetron has a significantly different chemical structure compared to other 5-HT₃ receptor antagonists. The classical 5-HT₃ receptor antagonists are derived from an indole moiety that mimics the structure of serotonin (Figure 2).^29,30 In contrast, palonosetron does not resemble serotonin; rather, it has a three-member ring moiety bound to a quinuclidine ring.^31,32 These unique structural differences of palonosetron could enable molecular interactions with the receptor different from those of the other 5-HT₃ antagonists.³⁰

Different molecular interactions between palonosetron and the other 5-HT₃ receptor antagonists are reflected in their pharmacology and can be best studied and understood by considering their binding curves. Ondansetron and granisetron engage in competitive antagonism of the 5-HT₃ receptor;³⁰ these 5-HT₃ receptor antagonists compete directly with serotonin for the same binding site on the 5-HT₃ receptor as the endogenous ligand. Hence, serotonin is released from the receptor upon ondansetron or granisetron binding. This direct competitive effect is visually represented as a line on a graph of the concentration of the antagonist needed to displace a bound ligand from the receptor (Figure 3).

In contrast, the antagonism of palonosetron is due to allosteric inhibition of the receptor.³⁰ Moreover, the compound binds to a site different than the binding site for serotonin. Binding of the antagonist affects the conformation of the receptor such that the endogenous ligand cannot bind or the receptor is not able to induce a downstream signal. Palonosetron binding to the 5-HT₃ receptor yields a curve that is typical of an allosteric inhibitor (Figure 3).³⁰

Furthermore, it is possible for a drug to bind to both the allosteric and competitive binding site. This enables the possibility of positive cooperativity, wherein the binding of the first molecule causes a conformational change to the structure of the receptor that enhances the likelihood of further antagonist molecules binding.³³ Indeed, structural studies have indicated that palonosetron exhibits positive cooperativity, in contrast to ondansetron and granisetron which exhibit simple bimolecular interactions (Figure 4).³⁰ Palonosetron binding triggers a conformational change in the receptor, which in turn increases the affinity between the receptor and other palonosetron molecules.

There could be several implications to the high affinity, allosteric, and positively cooperative binding properties of palonosetron. Palonosetron quite possibly binds more tightly and stays longer on the 5-HT₃ receptor. Also, these properties make palonosetron a much more efficient antagonist, since binding at one site makes it easier for another palonosetron molecule to bind. Lastly, the drug antagonist is much less likely to be displaced by the binding of serotonin than the other 5-HT₃ antagonists.³⁰ Since serotonin binding to and activation of the 5-HT₃ receptor is involved in emesis, it stands to reason that an allosteric antagonist that is harder to displace would be able to control emesis for longer periods of time.

Palonosetron has an extended half-life of approximately 40 hours³⁴—which is 4–10-fold longer than the half-lives of older 5-HT₃ receptor antagonists—and a 30-fold higher binding affinity for the 5-HT₃ receptor than other antagonists.²⁷ Moreover, a higher binding affinity may contribute to a long-lasting effect at the 5-HT₃ receptor.^30,31,35 In fact, both the plasma half-life and receptor-binding properties may be reflected in the extended duration of action of palonosetron.³⁶ The longer-lasting effects of the compound could be very important for minimizing the delayed emesis typically seen in PDNV. Of notable interest, other experiments have demonstrated that the effect of palonosetron in blocking serotonin signaling outlasts the presence of the drug on the receptor, in contrast to the other tested 5-HT₃ receptor antagonists.³⁰ Currently the mechanism behind this observation has not been elucidated; however, the most likely pharmacological interpretation is that the 5-HT₃ receptor has been internalized or desensitized due to palonosetron binding. With palonosetron binding, a cellular mechanism could be triggered that involves receptor internalization—a commonly observed phenomenon—and ultimately reduces the available serotonin receptors and signaling that can result in nausea and vomiting.³⁰ This hypothesis is currently being investigated as an explanation behind the lasting effects of palonosetron antagonism, compared to the shorter effects of the older 5-HT₃ compounds.³⁰

Conclusion

Emesis signaling is complicated by the sheer numbers of receptor pathways that are independently involved in producing nausea and vomiting. Nevertheless, this complexity has enabled the identification of a wide range of drug targets. For high-risk patients, it is particularly important to have multiple therapeutic options to enhance treatment effects. However, many of the receptor systems have important functions elsewhere in the body, and the blockage of their signaling can result in significant side effects.

Currently, there are no new classes of drugs to control nausea and vomiting in development. However, exciting new data indicate that novel drugs can be created which target existing receptors, but have sufficiently different pharmacological properties such that they may have different clinical behaviors as well.

References

Author notes