3.1 Introduction
Chemotherapy-induced nausea and vomiting (CINV) is a significant side effect of cancer therapy and can lead to poor compliance with therapy, treatment delays, dehydration, hospitalization, and a marked decrement in patient quality of life. With appropriate CINV control, safe outpatient administration of chemotherapy can be accomplished with no change in patients’ pre-therapy quality of life. Over the last 30 years, developments have improved in the control of CINV, including the advent of 5-hydroxytryptamine-3 (5-HT3) receptor antagonists which are an integral ingredient in regimens used today. With the variety of chemotherapy regimens today and the ongoing development of new combinations in addition to targeted therapy, there have been corresponding dynamic goals for control of not only CINV in general but in differentiating control of nausea over and above that of vomiting. In fact, current antiemetic therapy conceptually fulfills the true definition of “targeted therapy” as there is significant understanding of pathways involved in emesis as well as specific targeted antagonists to these pathways. This chapter will focus on the 5-HT3 pathway and the specific receptor antagonists to this pathway.
3.2 Physiology of CINV
The intrinsic emetogenicity of a given chemotherapeutic agent is the key determinant of the probability of clinical emesis, although chemotherapy dose, as well as patient factors such as female gender and young age may increase the probability of emesis [1]. Guideline groups divide chemotherapeutic agents into four emetogenic groups: high, moderate, low, and minimum [2] (Table 3.1). The clinical phases of emesis have been defined largely from observations using cisplatin where typically in the absence of antiemetics all patients will experience nausea and vomiting 1–2 h post administration. At 18–24 h the CINV abates, only to resurface and peak again at 48–72 h after the cisplatin administration [3]. As a result, acute emesis is defined as occurring within the first 24 h and delayed after 24 h (usually up to 120 h) [2]. Other agents may also cause delayed emesis, although not usually with the same biphasic pattern as cisplatin.
Table 3.1
Emetogenic potential of antineoplastic agents
Highly emetogenic (IV) | AC (doxorubicin or epirubicin with cyclophosphamide) | Dacarbazine | Ifosfamide >2 g/m2per dose |
Carmustine >250 mg/m2
| Doxorubicin >60 mg/m2
| Mechlorethamine | |
Cisplatin | Epirubicin >90 mg/m2
| Streptozocin | |
Cyclophosphamide >1,500 mg/m2
| |||
Moderately emetogenic (IV) | Aldesleukin >12–15 million IU/m2
| Clofarabine | Ifosfamide <2 g/m2per dose |
Amifostine >300 mg/m2
| Cyclophosphamide <1,500 mg/m2
| Interferon alfa >10 million IU/m2
| |
Arsenic trioxide | Cytarabine >200 mg/m2
| Irinotecan | |
Azacitidine | Dactinomycin | Melphalan | |
Bendamustine | Daunorubicin | Methotrexate >250mg/m2
| |
Busulfan | Doxorubicin <60 mg/m2
| Oxaliplatin | |
Carboplatin | Epirubicin <90 mg/m2
| Temozolomide | |
Carmustine < 250 mg/m2
| Idarubicin | ||
Low emetic risk (IV) | Ado-trastuzumab emtansine | Etoposide | Paclitaxel |
Amifostine <300 mg | 5-fluorouracil | Paclitaxel-albumin | |
Aldesleukin <12 million IU/m2
| Floxuridine | Pemetrexed | |
Brentuximab vedotin | Interferon alfa >5 <10 million IU/m2
| Pentostatin | |
Cabazitaxel | Ixabepilone | Pralatrexate | |
Carfilzomib | Methotrexate >50mg/m2 <250 mg/m2
| Romidepsin | |
Cytarabine 100–200mg/m2
| Mitomycin | Thiotepa | |
Docetaxel | Mitoxantrone | Topotecan | |
Doxorubicin (liposomal) | Omacetaxine | Ziv-aflibercept | |
Eribulin | |||
Minimal emetic risk (IV) | Alemtuzumab | Dexrazoxane | Pertuzumab |
Asparaginase | Fludarabine | Rituximab | |
Bevacizumab | Interferon alfa <5 million IU/m2
| Temsirolimus | |
Bleomycin | Ipilimumab | Trastuzumab | |
Bortezomib | Methotrexate <50 mg/m2
| Valrubicin | |
Cetuximab | Nelarabine | Vinblastine | |
Cladribine | Ofatumumab | Vincristine | |
Cytarabine <100 mg/m2
| Panitumumab | Vincristine (liposomal) | |
Decitabine | Pegaspargase | Vinorelbine | |
Denileukin diftitox | Peginterferon | ||
Moderate to high emetic risk (oral) | Altretamine | Estramustine | Procarbazine |
Busulfan >4 mg/day | Etoposide | Temozolomide >75mg/m2/day | |
Crizotinib | Lomustine (single day) | Vismodegib | |
Cyclophosphamide >100 mg/m2/day | Mitotane | ||
Minimal to low emetic risk (oral) | Axitinib | Gefitinib | Ruxolitinib |
Bexarotene | Hydroxyurea | Sorafenib | |
Bosutinib | Imatinib | Sunitinib | |
Busulfan <4 mg/day | Lapatinib | Temozolomide <75mg/m2/day | |
Cabozantinib | Lenalidomide | Thalidomide | |
Capecitabine | Melphalan | Thioguanine | |
Chlorambucil | Mercaptopurine | Topotecan | |
Cyclophosphamide <100 mg/m2/day | Methotrexate | Trametinib | |
Dasatinib | Nilotinib | Tretinoin | |
Dabrafenib | Pazopanib | Vandetanib | |
Erlotinib | Pomalidomide | Vemurafenib | |
Everolimus | Ponatinib | Vorinostat | |
Fludarabine | Regorafenib |
Advertisement
Animal studies have defined significant neuroanatomic components of the emetic reflex. Initially, Thumas proposed a single vomiting center (dorsal vagal nucleus) in 1891 based on canine studies [4]; Wang and Borison further refined the concept to include a “sensor” area in the area postrema (often called the chemoreceptor trigger zone) and an “effector” area, the vomiting center, in the medulla [5]. More recent studies suggest two areas of afferent input in the dorsal vagal nucleus, the area postrema and the nucleus tractus solitarius, and rather than a discrete “vomiting center,” several neuronal areas loosely organized to effect the emetic reflex, termed the “central pattern generator” [6].
The neurotransmitters involved in the emetic reflex related to chemotherapy are underpinned by 5-hydroxytryptamine (5-HT). Chemotherapy administration leads to local release of mediators by enterochromaffin cells in the proximal small intestine. These mediators include 5-HT, substance P, and cholecystokinin, which then bind to respective receptors (5-HT3, neurokinin-1, and cholecystokinin-1) located on terminal ends of abdominal vagal efferents [7]. This binding leads to a signal conducted by these vagal fibers which terminates in the nucleus tractus solitarius and leads to activation of the central pattern generator. The local release of 5-HT in the gastrointestinal tract and signal transduction by the vagal afferents is thought to be the chief mechanism whereby chemotherapeutic agents cause emesis [7]. Direct stimulation of the chemoreceptor trigger zone as described by Wang and Borison is a further mechanism, albeit thought to be a less important mechanism with CINV [8].
3.3 Development of First-Generation 5-HT3 Antagonists
In 1985, studies showed that high-dose metoclopramide (a dopamine-2 receptor antagonist) combined with dexamethasone provided meaningful protection from cisplatin-induced emesis at the expense of significant extrapyramidal side effects [9]. Curiously, at that time, results could not be replicated with other dopamine-2 receptor antagonists, suggesting the antiemetic effect was mediated by interaction with another neurotransmitter receptor. The 5-HT3 receptor emerged as a likely mediator of the antiemetic effect, and the pharmaceutical industry employed various strategies to develop selective 5-HT3 receptor antagonists. These included:
The structures of the four first-generation 5-HT3 antagonists are shown in Fig. 3.1 [14‐17]. Since ondansetron and granisetron are the most used agents with dolasetron no longer indicated for CINV and tropisetron in limited use, a greater focus will be devoted to ondansetron and granisetron. The role of 5-HT3 receptor antagonists in postoperative nausea and vomiting is beyond the scope of this text.
×
Advertisement
3.3.1 Ondansetron
Ondansetron is an indole derivative, selective 5-HT3 receptor antagonist with weak affinity for other 5-HT receptors and dopamine receptors. Oral formulations are rapidly absorbed with an approximate 60 % bioavailability [18]; ondansetron is extensively metabolized primarily by hydroxylation of the indole ring with subsequent conjugation [19]. The elimination half-life is approximately 4 h (Table 3.2); metabolites are not significant contributors to activity. The elderly and patients with hepatic impairment show reduced clearance; however, this is not clinically meaningful [19].
Table 3.2
5-HT3 receptor antagonist characteristics
5-HT3 receptor antagonist | Half-life (hours) | Oral bioavailability (%) | QTc prolongation | |
---|---|---|---|---|
First generation | Ondansetron | 4 | 59 | Yes |
Dolasetron (hydrodolasetron) | 7.3 | 76 | Yes | |
Tropisetron | 8 | 60 | Yes | |
Granisetron | 9 | 60 | Yes | |
Second generation | Palonosetron | 42 | 97 | No |
Adverse events with ondansetron are largely confined to headache and constipation. An EKG study conducted in 2012 demonstrated significant QTc prolongation with a 32 mg dose of ondansetron prompting the FDA to limit the dose administered intravenously to 16 mg [20].
3.3.2 Dolasetron
Dolasetron is a selective 5-HT3 receptor antagonist derived after extensive chemical substitutions to the cocaine molecule. Dolasetron is rapidly metabolized to the active metabolite hydroxydolasetron, which is predominantly excreted in the urine and has an elimination half-life of 7.3 h (Table 3.2) [21]. Renal impairment increases the elimination time, but is once again not clinically relevant.
As with ondansetron, headache and constipation are the most significant adverse effects. QTc prolongation is also evident at higher doses, and in 2010, due to the QTc concerns, the approval for intravenous dolasetron in CINV was withdrawn.
3.3.3 Tropisetron
Tropisetron is another selective competitive antagonist of the 5-HT3 receptor, derived by chemical modification of serotonin (5-HT) and thus is also an indole derivative. Oral absorption is rapid with approximately 60 % bioavailability; metabolism is similar to ondansetron (hydroxylation followed by conjugation), with metabolites being inactive [22]. In most patients, the elimination half-life is approximately 8 h (Table 3.2), although in some poor metabolizers, the half-life may be up to 45 h [23]. Tropisetron is not available in the USA, but is used in the East and Australia.
3.3.4 Granisetron
Granisetron is a selective 5-HT3 receptor antagonist derived by making chemical alterations to tropisetron. Oral bioavailability is similar to other agents and metabolism is the primary means of elimination. The elimination half-life is approximately 9 h (Table 3.2), with no significant changes noted in elderly and those with renal or hepatic impairment [24]. In addition to the intravenous and oral formulations, a transdermal delivery system (patch) is available in the USA, delivering granisetron directly through intact skin by passive diffusion with levels peaking at 48 h and sustained for further 5 days. Adverse events once again include headache and constipation and reports of QTc prolongation.
3.4 Development of Palonosetron
Chemical strategies that examined conformational alterations in 5-HT3 receptor antagonists led to compounds with significantly increased affinity for the receptor and further alterations led to the development of palonosetron, deemed a second-generation 5-HT3 antagonist for its enhanced receptor binding affinity and prolonged half-life (approximately 42 h) (Table 3.2) [25]. The structure of palonosetron is shown in Fig. 3.1 [26]. In addition, distinguishing palonosetron from first-generation agents, it does not appear to have a meaningful effect on QTc [27]. Oral bioavailability is excellent at 97 % [28].
3.5 Clinical Studies
Over the last 30 years, numerous clinical studies have been conducted in the CINV space that have established the benefits of first-generation 5-HT3 receptor antagonists as well as have helped to refine dosing and steroid combinations. During the same period, standards and definitions as far as end points and study design have also evolved, such that initial studies had more variability in terms of measured end points and use of patient-reported outcomes. Currently, important definitions include the acute phase (0–24 h post emetogenic chemotherapy), delayed phase (24–120 h post), complete response (no emesis and no use of rescue medication), and complete control (no emesis and no more than minimal nausea) [1]. While not perfect, these standard definitions allow studies to be interpreted in context and determine whether any differences seen are meaningful. In addition, standardization has assisted in the development of meaningful guidelines to help translate clinical trial benefits to global benefits.
Advertisement
3.5.1 Initial Studies
3.5.1.1 Placebo-Controlled Studies
Two studies compared the efficacy of ondansetron and granisetron to placebo in preventing cisplatin-induced emesis. Cubeddu [29] randomized 28 chemotherapy-naive patients about to receive cisplatin to ondansetron or placebo as antiemetic prophylaxis. The control of emesis was significantly improved in the active treatment arm in terms of number of emetic episodes, time to emesis, and need for rescue medication. Cupissol [30] similarly randomized 28 chemotherapy-naive patients to receive granisetron or placebo with their first dose of cisplatin. Once again, the active therapy group had significant control of emesis – 13 of 14 granisetron patients had no acute nausea or vomiting whereas only 1 of 14 placebo patients was free of nausea or vomiting in the first 24 h. In patients receiving cyclophosphamide (at that time deemed a moderately emetogenic agent irrespective of the combination), Cubeddu randomized 20 patients to ondansetron or placebo [31] and predictably 70 % of ondansetron-treated patients experienced no emesis, compared to 0 % in the placebo arm. Two further studies in Japan used a placebo design to confirm the activity of ondansetron and tropisetron [32, 33]. The continued use of placebo controls was strongly discouraged except for studies involving low emetogenic risk agents, and further studies needed to compare new treatments with the best available existing therapy [34]. An individual patient data meta-analysis [35] reinforced the dismal outcome in terms of emesis control in placebo-treated patients.
3.5.1.2 Dose-Finding Studies
Principles of cancer supportive care, unlike primary therapy, dictate that the lowest effective dose of an agent can be used rather than the maximally tolerated dose: many initial studies sought to define optimal dosing of the first-generation 5-HT3 receptor antagonists. In patients receiving cisplatin, a double-blind trial of three different ondansetron doses (0.015, 0.15, and 0.30 mg/kg, each given three times 4 h apart) demonstrated the 0.15 mg/kg dose was superior to the lower dose, and no added improvement with the higher dose was noted [36]. Granisetron was also evaluated in patients receiving cisplatin in a double-blind study [37]. Doses evaluated were 2, 10, and 40 μg/kg with the 10 and 40 μg/kg doses being superior to the lower dose in terms of preventing cisplatin-induced CINV. There was no difference observed between the 10 and 40 μg/kg doses. A second study examined granisetron doses of 40 μg/kg or 160 μg/kg and found no differences in efficacy [38]. Similarly studies with tropisetron and dolasetron [39, 40] confirmed intermediate doses as effective in preventing cisplatin-induced CINV with no added benefit to dose escalation. Importantly, though, even at much higher doses than required for maximum efficacy, adverse events were seldom significantly increased.
3.5.1.3 Comparison Studies with Older Agents
Ondansetron (0.15-mg/kg × 3 doses) was compared with high-dose metoclopramide (2 mg/kg × 6 doses) in a single-blind trial in 307 patients receiving high-dose cisplatin [41]. Ondansetron was found to be superior to metoclopramide and produced fewer adverse events; in particular no extrapyramidal effects were noted with ondansetron. Similar findings were seen with two earlier smaller studies; however, the schedule of ondansetron administration was not standard (continuous infusion) [42, 43].
Granisetron trials provided clues that helped the understanding of the neurophysiology of emesis and the pharmacology of the antiemetics: granisetron alone was superior to either chlorpromazine or prochlorpromazine given together with dexamethasone for moderately emetogenic chemotherapy [44, 45]; yet when granisetron alone was compared to high-dose metoclopramide with dexamethasone for high-dose cisplatin, no differences in antiemetic control could be discerned [45, 46].
These observations validated the original findings by Gralla using metoclopramide and confirming the role of the 5-HT3 receptor in emesis. In addition clues about the utility of corticosteroids emerged.
3.5.2 The Role of Corticosteroids
Corticosteroids, in particular dexamethasone, have been used as antiemetics for CINV and have demonstrated efficacy, but no clear mechanism of action since their protective effects appear much sooner than conventional corticosteroid mechanism would allow. With the advent of 5-HT3 receptor antagonists, numerous studies examined the question of single-agent 5-HT3 receptor antagonist compared to corticosteroid combinations. The studies are summarized by Jantunen [47] and combining 11 studies in a meta-analysis demonstrated the odds ratio of acute vomiting to be 0.42, strongly in favor of the combination arms. The Italian Group for Antiemetic Research went further to define the optimal dosing of dexamethasone with 5-HT3 receptor antagonists both for highly emetogenic [48] and moderately emetogenic therapy [49].
3.5.3 First-Generation 5-HT3 Antagonists Compared
The first study to compare ondansetron to granisetron randomized 496 patients to receive either ondansetron 8 mg, ondansetron 32 mg, or granisetron 3 mg prior to cisplatin-based therapy [50]. No significant difference was seen between any of the groups related to emesis, nausea, or adverse events. Multiple other studies compared various doses or schedules of ondansetron with granisetron both as prophylaxis for highly emetogenic chemotherapy and for moderately emetogenic chemotherapy [51‐56]. Other studies compared tropisetron [57, 58] as well as dolasetron [59‐61]. Despite numerous studies, no significant differences emerged. Even large meta-analyses did not show any appreciable differences in clinical efficacy and adverse events [62, 63], although in the larger analysis, tropisetron was not as effective as granisetron [63].
3.5.4 Intravenous Compared to Oral Therapy
Oral forms of 5-HT3 receptor antagonists were developed shortly after the intravenous forms and theoretically provide more convenient dosing. The agents are well absorbed and undergo some first-pass metabolism although they are generally conserved; the bioavailability of oral ondansetron is 59 % [18], granisetron 60 % [64], tropisetron 60 % [22], and dolasetron (hydrodolasetron) 76 % [65].
Initial studies of oral 5-HT3 receptor antagonists examined ranges of oral doses and efficacy was significantly superior to historical placebo controls for ondansetron [66], dolasetron [67], and granisetron [68]. A direct systematic comparison of the same agent, comparing intravenous and oral forms in large randomized studies was only performed in a single study. The Ondansetron Acute Emesis Study Group randomized 530 patients to receive a single dose of either oral ondansetron 24 mg or intravenous ondansetron 8 mg together with dexamethasone prior to cisplatin therapy [69]. The acute (<24 h) complete response rates (no emesis and no rescue medication) were 85 % and 83 % for the oral and intravenous groups, respectively. Further randomized studies compared different formulations and agents: oral granisetron was compared to intravenous ondansetron prior to highly emetogenic therapy [70] and prior to moderately emetogenic therapy [71]; oral ondansetron was compared to intravenous granisetron prior to cisplatin therapy [72]. All three studies showed no difference in emesis control between randomized arms.
Advertisement
Oral 5-HT3 receptor antagonists allowed the opportunity to explore the use of these agents beyond the first day in an attempt to lessen delayed emesis. Some studies showed reduced delayed emesis with the use of 5-HT3 receptor antagonists beyond 24 h [73], while others failed to show any benefit [74]. A meta-analysis of ten randomized studies, including five that used dexamethasone in the delayed setting and five that did not, determined the reduction in emesis risk with longer treatment [75]. The absolute emesis risk reduction for monotherapy was 8.2 % (95 % CI, 3.0–13.4 %), whereas the reduction was only 2.6 % (95 % CI, −0.6–5.8 %) with the use of dexamethasone.
Oral therapy also allowed the evaluation of these agents in other oncology settings, particularly in preventing radiation-induced emesis. In patients receiving fractionated upper abdominal radiation, a randomized study of 260 patients demonstrated significant control of radiation-induced emesis with granisetron 2 mg daily compared with placebo [76]. Spitzer et al. [77] compared oral granisetron or ondansetron prior to total body irradiation to historical control patients. Emesis rates were significantly reduced in patients receiving either 5-HT3 receptor antagonist compared to the historical controls.
3.5.5 Summary of Characteristics of First-Generation 5-HT3 Receptor Antagonists
The four first-generation 5-HT3 receptor antagonists discussed here were derived from different processes but essentially exhibit more similarities than differences. These characteristics are listed below and are used as principles to formulate guideline recommendations for CINV prophylaxis:
(i)
They all contain an indole ring and are highly selective antagonists of the 5-HT3 receptor.
(ii)
The elimination half-life is approximately 4–8 h.
(iii)
Oral formulations are well absorbed with an approximate 60 % bioavailability and are equivalent to intravenous formulations.
(iv)
The lowest effective dose has been determined.
(v)
Antiemetic efficacy is vastly superior to placebo for both highly and moderately emetogenic chemotherapy.
(vi)
Antiemetic efficacy is superior to older antiemetics except for high-dose metoclopramide.
(vii)
No significant differences in antiemetic efficacy are discernible between the four agents in individual studies.
(viii)
Antiemetic efficacy is superior when combined with corticosteroids.
(ix)
Side effects are largely confined to constipation and headache, with no appreciable increase in adverse events, even at escalated doses, although QTc prolongation has become a regulatory concern.
3.6 Palonosetron: A Second-Generation 5-HT3 Antagonist
Unlike the first-generation 5-HT3 antagonists, palonosetron is not based on an indole moiety; rather it contains a fused tricyclic ring and a quinuclidine moiety and has a half-life of approximately 40 h [28]. At the receptor level, it appears to bind to the 5-HT3 receptor more avidly than first-generation agents as well as exhibiting allosteric binding in contrast to the pure competitive binding seen with first-generation agents [78]. Further, it has been noted to cause receptor internalization [79], resulting in additional prolongation of duration of action. These unique properties are thought to account for some of the clinical efficacy differences seen in comparison to first-generation 5-HT3 receptor antagonists.
Advertisement
In the moderately emetogenic setting, palonosetron was compared to ondansetron [80] and dolasetron [81], respectively, in two double-blind randomized phase III trials. Both studies included three arms with only a single dose: palonosetron 0.25 mg and 0.75 mg and first-generation 5-HT3 receptor antagonist; no dexamethasone was administered as part of the protocol. Results from both studies demonstrated that a single dose of palonosetron (0.25 mg) was as effective as a single dose of a first-generation 5-HT3 receptor antagonist in preventing acute CINV and superior in preventing delayed CINV. In a randomized study conducted in the highly emetogenic setting, palonosetron was as effective as ondansetron and in a subset of patients also treated with dexamethasone appeared more effective [82].
Since the palonosetron registration studies were conducted without mandated corticosteroid therapy, the magnitude of benefit of palonosetron over the first-generation 5-HT3 receptor antagonists remained in question. A randomized study comparing palonosetron with dexamethasone to granisetron with dexamethasone in patients receiving highly or moderately emetogenic therapy helped to define the benefit [83]. There was no difference in the acute phase, but in the delayed phase, approximately 12 % more patients were emesis-free in the palonosetron arm as compared to the granisetron arm for both highly and moderately emetogenic therapy. Similar to first-generation 5-HT3 receptor antagonists, oral palonosetron has high bioavailability and similar efficacy to intravenous palonosetron with 0.5 mg being the preferred oral dose [84].
The improved efficacy observed with palonosetron has been attributed to the receptor-binding effects discussed above rather than the longer half-life, since administration of 5-HT3 receptor antagonists beyond day 1 results in negligible benefit in emesis control [75].
3.7 Novel Delivery Methods
Alternative routes of administration of first-generation 5-HT3 receptor antagonists have been explored, including subcutaneous, intramuscular, rectal, transdermal, and nasal/buccal sprays. In general, bioavailability of any route has been high with comparable efficacy where studied [85, 86]. Two delivery methods warrant further discussion: transdermal granisetron and polymer encapsulated slow-release granisetron.
3.7.1 Transdermal Granisetron
A transdermal formulation of granisetron is approved in the USA for the prevention of nausea and vomiting in patients receiving highly or moderately emetogenic chemotherapy. This granisetron transdermal delivery system is a 52 cm2 patch containing 34.3 mg of granisetron, which is delivered transdermally as 3.1 mg/24 h and essentially achieves a similar exposure to that of a 2 mg oral dose providing continuous delivery of granisetron over 6 days [87]. A randomized, double-blind study included 641 patients receiving chemotherapy and demonstrated that the transdermal delivery system was non-inferior to oral granisetron [88]. Although balanced between treatment arms, the patient population studied was heterogeneous in terms of emetogenicity of chemotherapy (high or moderate), prior chemotherapy exposure and the use of corticosteroids. In addition, a strong limitation on the utility of this delivery system is the requirement to place patch 24 h before scheduled chemotherapy. Finally, since the benefit of first-generation 5-HT3 receptor antagonists beyond the first day of chemotherapy is limited [75], protracted delivery would appear to hold little advantage. Intuitively, such delivery systems may be more useful for preventing emesis from radiation or oral agents: no such studies have been conducted. Nevertheless, the transdermal system is approved in the USA and remains a potential choice for prevention of chemotherapy-induced nausea and vomiting.
3.7.2 Sustained-Release Subcutaneous Granisetron
APF530 is a subcutaneously administered polymeric formulation of granisetron that provides slow, controlled, and sustained release of granisetron [89]. APF530 comprises 2 % granisetron and a polymer that is designed to undergo controlled hydrolysis, imparting the drug release characteristics. In a phase 3 non-inferiority trial, the clinical efficacy of APF530 250 mg subcutaneously and 500 mg subcutaneously (containing granisetron 5 and 10 mg, respectively) was compared with 0.25 mg palonosetron intravenously in patients receiving moderately or highly emetogenic chemotherapy [90]. Patients were stratified according to emetogenicity of chemotherapy and received study drug together with appropriate placebo (subcutaneous saline for palonosetron group; intravenous saline for APF530 group). All patients received guideline-appropriate doses of dexamethasone. The study demonstrated non-inferiority of the 500 mg dose of APF530 compared with palonosetron in preventing CINV both in the acute and delayed setting. Since the classification of emetogenicity in the initial study design was based on the older Hesketh algorithm [91], a reanalysis of the study was undertaken, using the latest American Society of Clinical Oncology emetogenic classification [92]. The reanalysis confirmed the initial findings of non-inferiority of APF530 [93].
The observations from the single study of APF530 question the mechanism used to explain the superiority of palonosetron over first-generation 5-HT3 receptor antagonists: that receptor binding was more important than the extended half-life. A clear explanation to unify these observations will require further study.
3.8 Conclusion
First-generation 5-HT3 receptor antagonists dramatically altered the delivery of cytotoxic chemotherapy, changing intolerable regimens to tolerable ones, shifting many chemotherapy regimens to the ambulatory setting, and improving quality of life for many patients. Further understandings of the mechanisms of emesis and clinical trial observations have allowed refinements in their use; 5-HT3 receptor antagonists form the backbone of most antiemetic regimens. Improvements have also been seen with palonosetron and with newer agents such as neurokinin-1 receptor inhibitors; however, nausea remains a persistent problem and will require further refinements in the use of multiple agents, together with a better understanding of the mechanisms of chemotherapy-induced nausea to improve overall CINV control.