John D. Hainsworth
SUMMARY OF KEY POINTS
|
Incidence
Etiology of Complication
Evaluation of the Patient
Grading of Complication
Treatment
|
INTRODUCTION
Nausea and vomiting are common side effects associated with systemic chemotherapy, and are among the adverse effects most feared by patients. [1] [2] Although these complications of treatment are usually self-limiting and seldom life-threatening, the deleterious effects on nutritional status and quality of life can be substantial. Many of the recently introduced targeted agents have less potential to produce nausea and vomiting. However, intensive combination chemotherapy continues to be the cornerstone of treatment for many types of cancer, ensuring that antiemetic therapy will continue to be an integral aspect of supportive care.
Antiemetic therapy has improved dramatically during the last 20 years. With optimum treatment, most patients receiving highly emetogenic chemotherapy do not experience any nausea or vomiting during the 24 hours after treatment. [3] [4] [5] [6] [7] [8] [9] [10] [11] However, delayed symptoms are more common and are often underestimated by treating physicians and nurses.[12] Accurate assessment of delayed nausea and emesis is essential in providing maximal intervention with recently available agents.
The identification of potent new antiemetics has been made possible by an improved understanding of the physiology of the emetic reflex. Critical assessment of the optimal use of new agents for patients receiving chemotherapy has been facilitated by the development of reproducible methods of assessing nausea and vomiting, and by the conduct of carefully designed, randomized clinical trials.
PHYSIOLOGY OF THE VOMITING REFLEX
The pioneering work of Borison and Wang[13] more than 40 years ago provided the basis for understanding the vomiting reflex. In studies using ablative techniques and electrical stimulation with microelectrodes (primarily in decerebrate cats), these investigators proposed the existence of two distinct sites in the brain stem believed to be critical for the control of emesis. The first of the sites, the so-called vomiting center, was thought to be located in the lateral reticular formation of the medulla. Electrical stimulation of this site triggered the vomiting reflex, whereas ablation prevented the vomiting induced by a variety of stimuli. The vomiting center was thought to be located adjacent to the other structures involved in the coordination of vomiting, including the respiratory, vasomotor, and salivary centers, and cranial nerves VIII and X. More recent studies have suggested that the “vomiting center” is actually not anatomically discrete but that the initiation of the vomiting reflex is controlled by a complex system of networks located in the nucleus tractus solitarius. [14] [15] The networks in this area control complex patterns of motor activity such as the vomiting reflex and are more accurately described as “central pattern generators.”
The second important center identified by Borison and Wang was the chemoreceptor trigger zone (CTZ), located in the area postrema at the ventral aspect of the fourth ventricle. This center, located outside the blood-brain barrier, is exposed to various noxious agents borne in the blood or cerebrospinal fluid. Although electrical stimulation of the CTZ does not produce vomiting, intimate connections to the vomiting center permit stimulation of this center after exposure to blood-borne toxins. Ablation of the CTZ abolishes vomiting induced by these agents.
Although these concepts have been retained and are integral to the current understanding of the vomiting reflex, several other important components have also been recognized. Input from the gastrointestinal tract, predominantly through afferent vagal fibers, is critical in initiating the vomiting reflex after ingestion of noxious substances.[16] Incoming vagal afferents connect with the vomiting center directly; an intact CTZ is not essential when vomiting is initiated by this mechanism. It is now known that in addition to ingested substances, some blood-borne substances, including chemotherapeutic agents, can trigger the vomiting reflex through activation of the vagal afferent mechanism.
Two additional components of this complex system involve the vestibular apparatus and the higher brain stem and cortical structures. The vestibular system is involved primarily in initiating the vomiting reflex in motion sickness. Input from higher cortical centers seems to be critical in a variety of conditions, including anticipatory emesis seen for patients who have previously experienced chemotherapy-induced emesis. The various components of the vomiting reflex are illustrated diagrammatically in Figure 39-1 , along with the clinical situations in which they are operative. Given the complexity of this system, it is not surprising that different pharmacologic approaches are necessary to control vomiting of different etiologies. Improved understanding of the neurochemistry of the emetic reflex has been important in developing antiemetics with new mechanisms of action. The initial focus of such investigation was the area postrema, where receptors for a large number of neuroactive agents have been identified. [17] [18] [19] Many of these neurotransmitters (e.g., dopamine, histamine, acetylcholine, norepinephrine, substance P) are in themselves emetogenic agents. The development of pharmacologic agents that block specific sets of receptors (e.g., dopamine, neurokinin-1) has resulted in the identification of valuable antiemetics, and it is likely that continued efforts in this area will be productive of additional valuable agents in the future.
|
Figure 39-1 Schematic diagram of the various pathways for initiation of the vomiting reflex. Clinical syndromes mediated by each mechanism are illustrated. |
In addition to neurotransmitters located in the CTZ, type 3 serotonin (5-hydroxytryptamine 3 or 5-HT3) receptors are present in large quantities on vagal and splanchnic afferents within the gastrointestinal tract.[20] These peripheral receptors are pivotal in the initiation of the acute nausea and vomiting caused by cisplatin and other strongly emetogenic chemotherapeutic agents; inhibition of this pathway by specific 5-HT3 receptor antagonists results in highly effective antiemetic therapy.[21]
CLINICAL FEATURES OF CHEMOTHERAPY-INDUCED EMESIS
Clinical Syndromes
Chemotherapy-induced nausea and vomiting can be subdivided into three distinct clinical syndromes, each having specific therapeutic implications. These syndromes and their clinical correlates are defined here; treatment approaches are considered later in the chapter. Because nausea and vomiting are common symptoms among patients with cancer, etiologies other than chemotherapy should also be considered. Among the diverse causes of nausea and vomiting among patients with cancer are intestinal obstruction, liver metastases, central nervous system involvement, and other medications (particularly narcotic analgesics). These etiologies should be considered especially when the time course or duration of nausea and vomiting is unusual for the known chemotherapy-induced syndromes.
Acute Nausea and Vomiting
Acute nausea and vomiting after the administration of chemotherapy occur within 24 hours after the chemotherapy dose. The nausea and vomiting during this phase are the most severe, hence the emphasis on therapeutic intervention during this phase. With most chemotherapeutic agents, acute nausea and vomiting begin 1 to 2 hours after intravenous administration. This delay in onset argues against a direct effect at the CTZ, which would be expected to produce emesis within minutes of intravenous drug administration. A peripherally mediated vomiting reflex, probably serotonin release from small intestinal mucosa, offers a better explanation of the delayed onset of emesis.[22] The onset of nausea and vomiting after the intravenous administration of cyclophosphamide is delayed even longer than with other agents, typically occurring 9 to 18 hours after administration of the drug.[23] The mechanism of cyclophosphamide-induced nausea and vomiting is unclear; the difference in the time of onset suggests that the mechanism might differ from that of other agents.
Delayed Nausea and Vomiting
Delayed nausea and vomiting occur 24 or more hours after chemotherapy administration. Although the severity is decreased in comparison with acute nausea and vomiting, the course can be more protracted, resulting in significant difficulties with hydration, nutrition, and performance status. Delayed emesis is most severe and frequent after administration of high-dose cisplatin; most patients treated with this drug experience some degree of delayed emesis, with onset most frequently 24 to 72 hours after chemotherapy.[24] In some patients onset can occur as late as 4 to 5 days after treatment, persisting for several days. Patients who have poor control of acute nausea and vomiting are more likely to experience delayed nausea and vomiting as well; however, delayed emesis can occur among patients who have complete emetic control during the first 24 hours after administration of chemotherapy.
The pathophysiology of delayed emesis remains unclear, but it seems likely that this syndrome is mediated centrally by different neurotransmitters. 5-HT3 receptor antagonists, which are highly effective in the prevention of acute emesis, have less activity in the treatment of delayed emesis. Conversely, the neurokinin-1 (NK-1) receptor antagonists, which block the action of substance P, have consistently shown activity against delayed emesis. Peripheral factors, including residual metabolites of chemotherapeutic agents or gastrointestinal mucosal damage, might also play a role.
Anticipatory Nausea and Vomiting
Anticipatory nausea and vomiting often occur among patients who have experienced poor control of emesis during previous courses of chemotherapy.[25] The onset can occur before or during chemotherapy administration. Because this is a conditioned response, certain associations with chemotherapy administration, such as the hospital environment or the oncologist's office, might trigger the onset of emesis.
Prognostic Factors
Multiple clinical factors that are important in determining the incidence and severity of chemotherapy-induced nausea and vomiting have been identified. These factors include the type of chemotherapy administered, certain patient characteristics, and the antiemetic regimen employed ( Table 39-1 ).
Table 39-1 -- Determinants of Chemotherapy-Induced Nausea and Vomiting
|
|||||||||||||||
|
|||||||||||||||
|
Chemotherapeutic Agents
The commonly used chemotherapeutic drugs are separated into five groups according to emetic potential in Table 39-2 . Drugs in category 5 produce emesis in greater than 90% of patients, whereas drugs in category 1 produce emesis in fewer than 10%. The drugs that cause emesis most frequently also cause the most severe emesis. Emesis is most severe during the first 8 hours after onset, but with strongly emetogenic drugs patients are often ill throughout the 24-hour period after administration.
Table 39-2 -- Acute Nausea and Vomiting without Prophylactic Treatment with Commonly Used Chemotherapeutic Agents
|
Level |
Frequency of Emesis (%) |
Agent |
|
5 |
>90 |
Carmustine >250 mg/m2 |
|
Cisplatin ≥50 mg/m2 |
||
|
Cyclophosphamide >1500 mg/m2 |
||
|
Dacarbazine |
||
|
Mechlorethamine |
||
|
Streptozocin |
||
|
4 |
60–90 |
Carboplatin |
|
Carmustine ≤250 mg/m2 |
||
|
Cisplatin <50 mg/m2 |
||
|
Cyclophosphamide >750 mg/m2 |
||
|
≤1500 mg/m2 |
||
|
Cytarabine >1 g/m2 |
||
|
Doxorubicin >60 mg/m2 |
||
|
Epirubicin >90 mg/m2 |
||
|
Melphalan >50 mg/m2 |
||
|
Methotrexate >1000 mg/m2 |
||
|
Procarbazine (oral) |
||
|
3 |
30–60 |
Arsenic trioxide |
|
Cyclophosphamide ≤750 mg/m2 |
||
|
Cyclophosphamide (PO) |
||
|
Doxorubicin 20–60 mg/m2 |
||
|
Epirubicin ≤90 mg/m2 |
||
|
Hexamethylmelamine (PO) |
||
|
Idarubicin |
||
|
Ifosfamide |
||
|
Irinotecan |
||
|
Lomustine |
||
|
Methotrexate 250–1000 mg/m2 |
||
|
Mitoxantrone |
||
|
Oxaliplatin >75 mg/m2 |
||
|
2 |
10–30 |
Cytarabine 100–200 mg/m2 |
|
Capecitabine |
||
|
Cetuximab |
||
|
Docetaxel |
||
|
Doxorubicin (liposomal) |
||
|
Etoposide |
||
|
5-Fluorouracil <1000 mg/m2 |
||
|
Gemcitabine |
||
|
90Y ibritumomab tiuxetan |
||
|
Methotrexate >50 mg/m2 <250 mg/m2 |
||
|
Mitomycin |
||
|
Paclitaxel |
||
|
Panitumomab |
||
|
Sorafenib |
||
|
Sunitinib |
||
|
Temozolomide |
||
|
Topotecan |
||
|
Tositumomab/iodine-131 tositumomab |
||
|
1 |
<10 |
Alemtuzumab |
|
Asparaginase |
||
|
Bevacizumab |
||
|
Bleomycin |
||
|
Bortezomib |
||
|
Chlorambucil (PO) |
||
|
Cladribine |
||
|
Denileukin diftitox |
||
|
Erlotinib |
||
|
Fludarabine |
||
|
Gemtuzumab ozogamicin |
||
|
Hydroxyurea |
||
|
Imitanib |
||
|
Melphalan (low dose, oral dosing) |
||
|
Methotrexate ≤50 mg/m2 |
||
|
Pentostatin |
||
|
Rituximab |
||
|
Thioguanine (PO) |
||
|
Trastuzumab |
||
|
Vinblastine |
||
|
Vincristine |
||
|
Vinorelbine |
Adapted from Hesketh PJ, Kris MG, Grunberg SM: Proposal for classifying the acute emetogenicity of cancer chemotherapy. J Clin Oncol 1997;15:103–109.
In general, the potential for acute nausea and vomiting increases with the dose of chemotherapy. Schedule of administration is also important with certain agents: Large intravenous bolus doses, or doses administered intravenously over a short period of time, are more likely to cause emesis than are smaller divided doses or continuous infusion.
The use of chemotherapeutic agents in combination increases the emetogenic potential of a treatment regimen. On the basis of the information in Table 39-2 , Hesketh and colleagues[26] have proposed a model for predicting the emetogenic potential of a combination regimen ( Table 39-3 ). As new combination regimens are introduced into clinical practice, application of this algorithm can result in optimum, cost-effective antiemetic therapy.
Table 39-3 -- Formula for Estimation of Emetogenic Potential of Combination Chemotherapy Regimens
|
|
Regimen |
Emetogenic Level of Agents in Regimen |
Emetogenic Level of Regimen |
|
CMF |
3 + 2 + 1 |
4 |
|
CAF |
3 + 3 + 2 |
5 |
|
Paclitaxel/carboplatin |
2 + 4 |
5 |
|
CVP |
3 + 1 + 1 |
3 |
Adapted from Hesketh PJ, Kris MG, Grunberg SM: Proposal for classifying the acute emetogenicity of cancer chemotherapy. J Clin Oncol 1997;15:103.
|
A, doxorubicin; C, cyclophosphamide; F, 5-fluorouracil; M, methotrexate; P, prednisone; V, vincristine. |
Patient Characteristics
Several patient characteristics are important predictors of the development and severity of acute chemotherapy-induced nausea and vomiting. These are age, gender, history of alcohol intake, and history of previous chemotherapy.
AGE.
Data are conflicting regarding the effect of patient age on the severity of chemotherapy-induced nausea and vomiting. However, increasing evidence indicates that chemotherapy-induced emesis occurs more frequently in younger patients. [21] [22] Fortunately, the best current antiemetic agents are effective and well tolerated by patients of all ages.
GENDER.
Data from large prospective studies indicate that females have more severe and frequent chemotherapy-induced nausea and vomiting than males, even after controlling for chemotherapy regimen. In one study all patients received cisplatin-containing regimens and were treated with ondansetron; more women receiving high-dose cisplatin with either 5-fluorouracil or etoposide for lung cancer or head and neck carcinomas had poor control of emesis than did men receiving these same regimens (49% vs. 29%, respectively).[27]
HISTORY OF ALCOHOL INTAKE.
Patients with a history of chronic alcohol intake (four to five mixed drinks per day) have more effective control of chemotherapy-induced nausea and vomiting when optimal antiemetics are used. [28] [29] In a prospective study of 52 patients receiving high-dose cisplatin along with combination antiemetic therapy, 93% of those with a high alcohol intake experienced no emesis, as opposed to 61% of patients without this history.[28] It is important to emphasize, however, that the administration of highly emetogenic chemotherapy to these patients, in the absence of appropriate antiemetic therapy, still results in a high incidence of severe acute nausea and vomiting. The mechanism of the alcohol effect is unclear; it is possible, however, that various receptor sites are less sensitive among patients with a history of alcohol intake and that blockade of these receptors is relatively easy with appropriate antiemetics.
PREVIOUS CHEMOTHERAPY.
Patients who have experienced poor control of emesis during previous chemotherapy are more likely to have unsatisfactory results with subsequent antiemetics.[27] The development of an anticipatory component to the nausea and vomiting is certainly one influence; whether additional influences are also involved is unknown.
Conduct and Interpretation of Clinical Antiemetic Trials
Because multiple patient characteristics and chemotherapy variables influence the incidence and severity of chemotherapy-associated nausea and vomiting, optimal design and interpretation of clinical trials with new antiemetics or new combination regimens requires that these factors be taken into account. Definitive demonstration of the superiority of an agent or regimen requires a large, randomized trial in which the comparison groups are matched with respect to the various patient characteristics and chemotherapy received. Patients should be receiving their initial dose of chemotherapy, so as to avoid the confounding effects of anticipatory emesis. Crossover trials are also difficult to interpret for this reason. Optimally, the treatment received should be double-blinded to avoid investigator or patient bias. Because effective regimens now exist for all subgroups of patients, new treatments should be compared with existing treatments; the inclusion of a “no-treatment” or placebo arm in a randomized trial is inappropriate. The completion of this type of large controlled randomized trial has hastened the improvement of antiemetic therapy and the rapid incorporation of new treatments into clinical practice.
When interpreting the results of antiemetic trials, special attention should also be given to the definitions of therapeutic response and the methods used to assess efficacy. Most trials measure the number of emetic episodes as the primary efficacy parameter, because this is an objective measurement. Some trials also measure various secondary efficacy parameters, including nausea, food intake, and overall patient satisfaction with treatment. Even when emesis is used to measure efficacy, comparisons between trials must be made with caution, as definitions of response have varied also. “Complete response,” in various studies, has been defined as “no vomiting,” “no vomiting and only mild nausea,” or “no vomiting and no nausea” during the 24 hours after chemotherapy. Clearly, response rates can vary substantially, depending on the definition used. Standardization of response assessment would greatly aid in the interpretation of future trials.
TREATMENT OF CHEMOTHERAPY-INDUCED NAUSEA AND VOMITING
Acute Nausea and Vomiting
Several families of drugs with antiemetic activities have been identified. Table 39-4 lists the classes of antiemetic agents in current use, in approximate order of antiemetic potency. Only the 5-HT3 receptor antagonists, the substituted benzamides, and the new class of NK-1 receptor antagonists show marked activity against highly emetogenic chemotherapy. Because no single agent is ideal, combination antiemetic regimens have been developed, which have further improved efficacy. Increased understanding of the mechanism of action of the various compounds has led to more rational development of combination regimens; in general, the most effective regimens use agents with different mechanisms of action.
Table 39-4 -- Antiemetic Agents: Recommended Dosing
|
|
RECOMMENDED DOSE |
|
|
Antiemetic Agent |
Acute Emesis |
Delayed Emesis |
|
5-HT3 Antagonists |
||
|
Ondansetron |
0.15 mg/kg or 8 mg IV; 12–16 mg PO |
8 mg PO bid × 2–3 days |
|
Granisetron |
1 mg IV or PO |
|
|
Dolasetron |
1.8 mg/kg or 100 mg IV; 100–200 mg PO |
|
|
Palonosetron |
0.25 mg IV |
|
|
Neurokinin-1 antagonists |
||
|
Aprepitant |
125 mg PO (or 115 mg IV) day 1, then 80 mg days 2, 3 |
|
|
Corticosteroids |
||
|
Dexamethasone |
8 mg IV or PO |
4–8 mg PO bid days 2–4 |
|
Dopamine antagonists |
||
|
Metoclopramide |
2–3 mg IV prechemotherapy, repeat 2 hr after chemotherapy |
20–40 mg PO qid days 2–5 |
|
Prochlorperazine |
10 mg PO or IV every 3–4 hrs as needed |
|
|
Haloperidol |
1–3 mg IV or PO every 2–6 hrs as needed |
|
|
Benzodiazepines |
||
|
Lorazepam |
1–2 mg IV every 4 hrs as needed |
|
|
Cannabinoids |
||
|
Dronabinol |
5–10 mg PO every 3–4 hrs as needed |
|
5-HT3 Receptor Antagonists
The selective 5-HT3 receptor antagonists are the most effective family of antiemetics in the treatment of acute emesis. These agents block the serotonin type 3 receptors and are thought to exert their antiemetic activity primarily through peripheral blockade in the small intestine (see the preceding discussion). Before 2003, three agents in this class were available in the United States: ondansetron, granisetron, and dolasetron. All three of these drugs proved superior to high-dose metoclopramide (the previous standard) in the prophylaxis of cisplatin-induced emesis, [30] [31] [32] [33] [34] [35] and all were superior to standard agents when used to prevent emesis associated with cyclophosphamide-based regimens. [36] [37] [38] Dose-ranging studies with all drugs demonstrated an efficacy plateau, [39] [40] [41] and multiple large randomized trials comparing these agents to one another showed equivalent efficacy. [3] [4] [5] [6] [7] [8] [9] [10] [11] For patients receiving high-dose cisplatin therapy, use of appropriate doses of any one of these three agents results in complete control of emesis in 50% to 70% of patients during the first 24 hours after administration of chemotherapy. Complete control of emesis ranges from 70% to 80% among patients receiving moderately emetogenic regimens, usually cyclophosphamide based ( Box 39-1 ).
Box 39-1
|
MANAGEMENT APPROACH |
|
Chemotherapy-Induced Nausea and Vomiting Antiemetic therapy for acute chemotherapy-induced nausea and vomiting should be based on the emetic potential of the chemotherapy regimen being used and should take into account individual patient risk factors (e.g., sex, age, history of alcohol use, previous emesis with chemotherapy). Excellent algorithms for the prediction of the likelihood of emesis with single chemotherapeutic agents or combination regimens have been developed. For optimal control of acute nausea and vomiting, patients receiving chemotherapy with moderate or high emetogenic potential should receive prophylaxis with a 5-HT3 receptor antagonist plus dexamethasone. Palonosetron may offer modest advantage over other 5-HT3 receptor antagonists, particularly in patients receiving highly emetogenic regimens. Patients receiving mildly emetogenic therapy should receive dexamethasone (8 mg intravenously or orally); a 5-HT3 receptor antagonist should be added with subsequent courses only if antiemetic control is inadequate. Patients with breakthrough nausea and vomiting should receive either prochlorperazine (10 mg orally or intravenously) or lorazepam (1 mg intravenously every 4 hours). Additional doses of 5-HT3 receptor antagonists or dexamethasone during the first 24 hours are usually ineffective. Delayed nausea and vomiting occurs more than 24 hours after chemotherapy is administered and should be anticipated among patients receiving highly emetogenic regimens (particularly cisplatin or cyclophosphamide). These patients should routinely receive prophylaxis: aprepitant should be added to the antiemetic regimen, and palonosetron is the 5-HT3 receptor antagonist of choice. |
In July 2003 a new 5-HT3 receptor antagonist, palonosetron, was approved by the U.S. Food and Drug Administration for the prophylaxis of acute and delayed chemotherapy-induced nausea and vomiting. Compared with other 5-HT3 receptor antagonists, palonosetron has a higher receptor binding affinity and a longer half-life (approximately 40 hours). In two phase III randomized clinical trials palonosetron provided superior control of acute and delayed nausea and vomiting when compared with either ondansetron or dolasetron. [42] [43] Both trials contained patients receiving moderately emetogenic chemotherapy, and most patients did not receive concurrent dexamethasone.
The adverse events produced by all available 5-HT3 receptor antagonists are essentially identical. Mild to moderate headache is the most frequently observed toxicity and is produced by all three drugs in approximately 20% to 30% of patients. Other adverse events, including constipation and diarrhea, are mild and uncommon.
Recommended intravenous and oral doses of the 5-HT3 receptor antagonist and other antiemetics are outlined in Table 39-4 . For prophylaxis of acute nausea and vomiting, a single dose during the first 24 hours is as effective as multiple doses. Therefore, the use of additional doses of these agents for “breakthrough” vomiting during this time is to be discouraged, and the writing of “as needed” orders for these agents should be avoided. Oral preparations are available for all 5-HT3 receptor antagonists except palonosetron and have shown equivalent efficacy. [44] [45] [46]
Neurokinin-1 Receptor Antagonists
The NK-1 receptor is a component of the centrally mediated vomiting reflex that mediates the emetogenic action of substance P, a tachykinin contained in vagal afferents innervating the area postrema and nucleus tractus solitarii in the brain stem.[19] Several selective NK-1 receptor antagonists have shown antiemetic activity, [47] [48] and one of these, aprepitant, is currently approved by the Food and Drug Administration for the prevention of acute and delayed chemotherapy-induced nausea and vomiting. [49] [50] [51] [52] The most substantial contribution of this drug has been in the management of delayed emesis following moderately or highly emetogenic chemotherapy. In the pivotal randomized phase III study, patients treated with high-dose cisplatin received prophylaxis with either granisetron/dexamethasone or granisetron/dexamethasone plus aprepitant.[49] Patients who received aprepitant more frequently had no vomiting during the first 24 hours (93% vs. 67%; P < 0.001) and also during days 2 through 5 (82% vs. 33%; P < 0.001). Similar results have subsequently been obtained in patients receiving cyclophosphamide-containing regimens.[52] Side effects with aprepitant are mild and infrequent, and include fatigue, constipation, hiccoughs, and headache.
Substituted Benzamides
High-dose metoclopramide was the first drug to demonstrate substantial antiemetic activity among patients treated with high doses of cisplatin.[53] The mechanism of action was initially attributed to dopamine receptor antagonism. However, high-dose metoclopramide also inhibits the 5-HT3 receptors, probably a more important mechanism.[54] Today, high-dose metoclopramide is rarely used, as a result of the higher efficacy and lower toxicity of the selective 5-HT3 receptor antagonists.
The major side effects associated with high-dose metoclopramide are extrapyramidal reactions caused by dopamine receptor antagonism. [53] [55] Acute dystonic reactions are the most dramatic adverse effects but are relatively uncommon in adults, occurring in only 2% to 5% of cases. Akathisia, though less dramatic, is a more common problem and often persists for several hours. Both adverse effects are more common in young patients. In most adults, these side effects are not difficult to control or prevent. Intravenous diphenhydramine quickly ends an acute dystonic reaction, and the addition of lorazepam to metoclopramide-containing regimens greatly reduces the incidence of extrapyramidal reactions.[56]
Corticosteroids
The antiemetic mechanism of action of the corticosteroids is unclear. Unlike most other antiemetics, there is no current evidence that neurotransmitter blockade is involved. The antiemetic activity of the corticosteroids has been confirmed in several trials, predominantly for patients receiving moderately emetogenic chemotherapy. [57] [58] Corticosteroids that have been used as antiemetics include dexamethasone, methylprednisolone, and occasionally prednisone. Although no obvious differences in efficacy have been demonstrated among the corticosteroids, dexamethasone is almost universally used currently. A recent randomized dose-finding study identified a single 8-mg dose of dexamethasone before treatment as optimal.[59] Because of their moderate antiemetic efficacy, corticosteroids should be used as single agents only in the prophylaxis of mildly emetogenic chemotherapy. However, they consistently add to the effect of other antiemetics when used in combination (see the discussion later in this chapter), presumably because they have a different mechanism of action.
The adverse effects of short courses of corticosteroid are mild and infrequent. Additional caution must be used in treating patients with diabetes mellitus or other conditions predisposing to difficulties with steroids. Occasional acute psychotic reactions have been observed.
Phenothiazines
The phenothiazines were the first family of agents to demonstrate substantial antiemetic activity and are thought to act primarily as antidopaminergic agents. Several of these agents, including prochlorperazine, promethazine, and thiethylperazine, are still used frequently. In 1963 a randomized, double-blind, placebo-controlled trial documented the superiority of prochlorperazine and thiopropazate to placebo in the control of nausea and vomiting induced by fluorouracil.[60] Efficacy has also been documented against other moderately emetogenic chemotherapeutic agents.[61] These agents are ineffective against highly emetogenic chemotherapy, however, and should currently be used only in combination with other more effective agents.[53] The side effects of the phenothiazines include sedation, akathisia, and, less commonly, acute dystonic reactions.
Benzodiazepines
Lorazepam is the only benzodiazepine that has found widespread use in antiemetic therapy. Direct antiemetic effects of lorazepam are minor; however, the sedative, anxiolytic, and amnesic effects have made this drug ideal for use in combination regimens.[62] Several trials have documented improved patient acceptance of lorazepam-containing combination regimens, even when the objective antiemetic efficacy increased only slightly.[56] The use of intravenous lorazepam often causes marked sedation lasting several hours, which limits its use in the outpatient setting. In addition, some patients experience confusion, amnesia, and transient enuresis.
Butyrophenones
The butyrophenones, haloperidol and droperidol, have antiemetic activity as a result of specific dopamine receptor blockade. [63] [64] Common side effects of the butyrophenones include sedation, dystonic reactions, and akathisia; in addition, hypotension is occasionally encountered. Because the butyrophenones are less efficacious than the 5-HT3 receptor antagonists and high-dose metoclopramide, their use should be reserved for the occasional patient who achieves poor results with standard antiemetic treatment.
Cannabinoids
Anecdotal reports of reduced emesis among patients smoking marijuana during chemotherapy stimulated interest in the cannabinoids during the early 1980s. Several cannabinoids have been evaluated as antiemetics; at present dronabinol is the only commercially available agent in this class. This drug has antiemetic activity for patients receiving moderately emetogenic chemotherapy; in this setting, it has been more effective than prochlorperazine.[65] The mechanism of action of the cannabinoids is incompletely defined; a central nervous system site of action has been postulated because of the marked psychoactive properties of these agents. Clinical use of cannabinoids is limited because of their unfavorable toxicity profile in some patients. Frequent toxicities include dysphoria, hallucinations, vertigo, dry mouth, sedation, and disorientation. These side effects are more common in elderly patients. The cannabinoids should not be considered for first-line antiemetic therapy, because more effective and better tolerated agents exist. They should be considered for the occasional patient with mild to moderate nausea and vomiting who has either poor tolerance or poor response to other antiemetics.[66]
Combination Antiemetic Therapy
Because none of the antiemetic agents is ideal when used alone, a variety of combination regimens has been developed in an attempt to improve efficacy. The most successful combinations use drugs with different mechanisms of action and nonoverlapping toxicities.
The corticosteroids have been most extensively evaluated as “second drugs” in combination regimens because of their ease of administration, minimal toxicity, and different mechanism of action. When combined with the 5-HT3 receptor antagonists, dexamethasone has consistently improved antiemetic efficacy among patients receiving moderately or highly emetogenic regimens. [67] [68] [69] In a recent meta-analysis of randomized trials comparing various regimens with or without dexamethasone, complete protection from emesis was increased by 16% when dexamethasone was added to a 5-HT3 receptor antagonist.[70] The addition of dexamethasone to high-dose metoclopramide also improves antiemetic efficacy.[71] A single dexamethasone dose of 8 mg before therapy is as effective as higher doses or multiple dose schedules.[59] Therefore, a corticosteroid should be included in the prophylactic treatment of all patients receiving moderately or highly emetogenic chemotherapy, unless specific contraindications exist.
Aprepitant is also an agent with proven efficacy in combination regimens, usually as a third drug (added to a 5-HT3/dexamethasone combination). The addition of aprepitant reduces acute and delayed emesis in patients receiving moderately or highly emetogenic chemotherapy. [49] [50] [51] [52] Aprepitant should be included in the prophylactic antiemetic treatment of all patients receiving highly emetogenic chemotherapy and should be considered in patients receiving moderately emetogenic chemotherapy who are at high risk for emesis (e.g. female, younger patients, emesis with previous chemotherapy). If aprepitant is not initially included in the antiemetic regimen, it should be added during subsequent cycles for patients who have poor control of emesis.
The addition of lorazepam to antiemetic regimens was common when high-dose metoclopramide was the major antiemetic agent, in large part as a result of the reduction of extrapyramidal side effects.[56]However, the contribution of lorazepam to current 5-HT3 receptor antagonist combinations is unclear, and sedation is consistently increased when lorazepam is added. The use of lorazepam should be considered only for patients with a high level of anxiety before treatment.
Anticipatory Nausea and Vomiting
Because anticipatory nausea and vomiting are conditioned responses, effective control of chemotherapy-induced nausea and vomiting prevents the development of this reflex and is therefore the best strategy for preventing this problem. [25] [72] The highly effective combination regimens in current use have greatly decreased the prevalence of this problem. For patients in whom anticipatory nausea and vomiting develop, treatment with anxiolytics such as the benzodiazepines is sometimes effective.[62] In addition, various nonpharmacologic approaches, including hypnosis and behavioral modification, have shown some benefit for these patients. [25] [73] [74]
Delayed Nausea and Vomiting
Until recently, control of delayed nausea and vomiting was poor as compared with the control achieved during the first 24 hours after chemotherapy. The combination of metoclopramide (0.5 mg/kg four times daily for 4 days) and dexamethasone (8 mg twice daily for 2 days, then 4 mg twice daily for 2 days) was superior to dexamethasone alone in patients receiving cisplatin-based chemotherapy (35% vs. 11% control rate).[75] However, most patients did not have adequate control of emesis, and side effects were bothersome for some patients. Similarly, the contribution of the initial 5-HT3 receptor antagonists was modest, when compared with their excellent activity in treating acute nausea and vomiting. Oral ondansetron (8 mg orally twice daily, days 2–6) had modest benefit versus placebo,[76] but in other trials the efficacy of a 5-HT3 antagonist was not any better than dexamethasone. [77] [78] Despite these modestly effective interventions, a recent study showed that the rates of delayed nausea (54%) and vomiting (32%) remained substantial in patients receiving moderately or highly emetogenic chemotherapy.[79] Furthermore, 23% of patients with complete control during the first 24 hours had either nausea or vomiting during the delayed phase.
With the introduction of palonosetron and aprepitant, the treatment of delayed nausea and vomiting has improved substantially. Palonosetron, the newest 5-HT3 receptor antagonist, is more effective than either ondansetron or dolasetron in controlling delayed nausea and vomiting. [42] [43] In patients receiving moderately emetogenic chemotherapy, a single dose of palonosetron (0.25 mg) provided complete control of delayed nausea and vomiting in 48%, as compared with 36% for dolasetron 100 mg.[42] Similar results occurred when palonosetron 0.25 mg was compared to ondansetron 32 mg (74% vs. 55% control of delayed emesis; P < 0.001).[43] The differences between palonosetron and the other 5-HT3 receptor antagonists are unexplained but may relate to the longer half-life of palonosetron or to the improved control of acute emesis.
The addition of aprepitant, the first NK-1 antagonist, has been even more important in controlling delayed nausea and vomiting. [49] [50] [51] [52] When added to granisetron/dexamethasone in patients receiving cisplatin-based chemotherapy, aprepitant reduced the incidence of delayed emesis from 67% to 18% (P < 0.001).[49] Improvement also occurred when aprepitant was added to ondansetron/dexamethasone in patients receiving cyclophosphamide/anthracycline chemotherapy (complete control improved from 42.5% to 50.8%; P = 0.019).[52]
Currently, optimal management of delayed chemotherapy-induced nausea and vomiting requires prediction of high-risk patients and administration of the most effective antiemetic regimen, beginning with the first cycle of chemotherapy. High-risk patients include all patients receiving highly emetogenic chemotherapy regimens, and certain subgroups (e.g. young patients, females) receiving moderately emetogenic regimens. In these patients, palonosetron should be the 5-HT3 receptor antagonist of choice, and aprepitant should be added to the antiemetic regimen. Similar substitutions or additions should be made in other patients who experience delayed nausea or vomiting regardless of the chemotherapy regimen they are receiving.
Radiation-Induced Nausea and Vomiting
Radiation-induced nausea and vomiting are common with some types of radiation therapy and are related to the size of the radiation portal, the dose delivered, and the site of radiation. Radiation-induced emesis occurs acutely in more than 90% of patients receiving total-body irradiation. Among patients receiving conventional daily doses of radiotherapy (2 Gy/fraction), emesis develops within 2 to 3 weeks in about 50% of patients receiving an upper abdominal portal.[80] The mechanism of radiation-induced emesis remains unclear, but release of serotonin from the gastrointestinal enterochromaffin cells and subsequent involvement of the gastrointestinal 5-HT3 receptors and vagal afferent fibers is most likely.
Most of the antiemetic agents found active against chemotherapy-induced nausea and vomiting also have some activity against radiation-induced emesis; however, few randomized trials have been performed to identify an optimal regimen. Dexamethasone or ondansetron, used as single agents, were superior to placebo in patients receiving upper abdominal irradiation [81] [82] or total-body irradiation.[83] With multifractionated courses, the use of daily dexamethasone is problematic; however, the addition of dexamethasone (4 mg orally daily, during the first 5 days only) to ondansetron (8 mg orally twice daily) improved results in patients receiving upper abdominal radiation therapy.[84]
Given the postulated mechanism of radiation-induced emesis and the available clinical data, it seems likely that the 5-HT3 receptor antagonists are the most active antiemetic agents in this setting. Routine prophylaxis for acute and delayed nausea and vomiting should accompany total-body irradiation. For patients receiving upper abdominal irradiation, in which the incidence of nausea and vomiting is lower, initial prophylaxis with phenothiazines or other less expensive agents might be a reasonable option. However, these patients should receive daily oral prophylaxis with a 5-HT3 receptor antagonist if emesis is uncontrolled with other agents; use of concurrent dexamethasone for the first 5 days should also be considered.
SUMMARY OF RECOMMENDATIONS FOR COMBINATION ANTIEMETIC THERAPY
Optimal selection of antiemetic therapy is based on the emetic potential of the planned chemotherapy regimen. The information presented in Tables 39-2 and 39-3 [2] [3] should be used to assign an emetic level to the chemotherapy agent or regimen being used. Optimal doses of all antiemetic agents are contained in Table 39-4 .
All patients receiving regimens with emetic level 4 or 5 should receive routine prophylaxis for acute nausea and vomiting with a 5-HT3 receptor antagonist plus dexamethasone. Because all of these patients also have substantial risk for delayed emesis, aprepitant should be added, and palonosetron should be considered the 5-HT3 antagonist of choice. Patients receiving level 3 regimens are usually well controlled with a 5-HT3 receptor antagonist plus dexamethasone; routine prophylaxis for delayed emesis is not necessary. Patients receiving mildly emetogenic regimens (category 1 or 2) do not require routine use of a 5-HT3 receptor antagonist and are usually managed effectively with either dexamethasone alone (category 2) or no prophylactic antiemetic therapy (category 1).
FUTURE DIRECTIONS
Major improvements in antiemetic therapy during the past 20 years have resulted in complete protection from chemotherapy-induced nausea and vomiting in most patients receiving chemotherapy. A few patients (fewer than 20%) continue to have problems during the first 24 hours after treatment, and in most of these patients breakthrough nausea and/or vomiting occurs more than 16 hours after administration of chemotherapy. Delayed nausea and vomiting remain common, but therapy has improved markedly with the introduction of aprepitant and palonosetron.
The evolution of standard chemotherapy for many types of cancer has had an impact on the antiemetic therapy required. Two highly emetogenic chemotherapeutic agents, mechlorethamine and dacarbazine, are now uncommon in clinical practice. More importantly, the use of cisplatin has declined greatly during the last several years mostly because of its replacement by carboplatin, a drug with a lower emetogenic potential. Many of the more recent cytotoxic agents (e.g., paclitaxel, docetaxel, gemcitabine, topotecan, vinorelbine) are mildly emetogenic agents. Increasing use of these new drugs has also fortuitously allowed the more effective control of chemotherapy-induced emesis. Finally, it is likely that future targeted agents, many of which will be administered orally, will not be associated with nausea or vomiting in most patients. Recent experience with several monoclonal antibodies (e.g., rituximab, trastuzumab, bevacizumab) and oral tyrosine kinase inhibitors (e.g., imatinib, erlotinib) con-firms the low frequency of nausea, even in the absence of any prophylaxis.
Future improvement in antiemetic therapy will require continued improvement in understanding of this complex physiologic reflex, and identification of new mediators. Continued well-designed clinical trials are necessary to define the most effective—and the most cost-effective—ways of incorporating new agents into current antiemetic regimens.
REFERENCES