Olanzapine Administration Reduces Chemotherapy-Induced Nausea Behavior in Rats

Abstract

Nausea and vomiting are consistently identified among the most distressing side effects of chemotherapy. In recent years, Olanzapine (OLZ) treatment was added to anti-emetic guidelines as a treatment for chemotherapy-induced nausea and vomiting (CINV), despite little available data supporting a mechanism behind the positive benefits of the drug. Here, we examine whether OLZ reduces cisplatin chemotherapy-induced side effects on food intake and pica behavior in rats (i.e., kaolin intake, a proxy for nausea/emesis). Behavioral experiments tested whether systemic or hindbrain administration of OLZ ameliorated cisplatin-induced pica, anorexia, and body weight loss in rats. We also tested whether systemic OLZ reduces cisplatin-induced neuronal activation in the dorsal vagal complex (DVC), a hindbrain region controlling emesis. Lastly, given their role in regulating feeding and emesis, circulating ghrelin levels and central mRNA expression levels of serotonin (HT) receptor subunits, including 5-HT2C, were measured in brain regions that regulate CINV and energy balance in an exploratory analysis to investigate potential mediators of OLZ action. Our results show that both systemic and hindbrain administration of OLZ attenuated cisplatin-induced kaolin intake and body weight loss, but not anorexia. Systemic OLZ decreased cisplatin-induced c-Fos immunofluorescence in the DVC and prevented cisplatin-induced reductions in circulating ghrelin levels. IP OLZ also blocked cisplatin-induced increases in Htr2c expression in DVC and hypothalamic micropunches. These data suggest hindbrain exposure to OLZ is sufficient to induce reductions in cisplatin-induced pica and that central serotonergic signaling, via 5-HT2C, and changes in circulating ghrelin may be potential mediators of olanzapine anti-emetic action.

Every year, more than 1.6 million people in the United States are diagnosed with cancer, a third of which undergo chemotherapy (Hesketh, 2008). Although chemotherapy may be critical for survival, chemotherapeutic agents produce severe side effects which decrease quality of life and can lead to treatment attrition or discontinuation (Griffin et al., 1996; Molassiotis et al., 2008). Cisplatin is one of the most potent and widely studied emetogenic antineoplastic drugs (Hesketh, 2008; Kris et al., 1985) and produces severe nausea, vomiting, anorexia, and cachexia in a variety of species, including humans (Andrews & Horn, 2006; Horn et al., 2013). Current anti-emetic pharmacotherapies have improved control of chemotherapy-induced vomiting (Aapro et al., 2015; Hesketh et al., 2010). However, these drugs have been less effective in the management of chemotherapy-induced nausea (Feyer & Jordan, 2011; Olver et al., 2014). This lack of control is especially prominent in the prevention of “delayed” chemotherapy-induced nausea and vomiting (CINV; i.e. nausea and vomiting occurring more than 24 hr after chemotherapy; Feyer & Jordan, 2011) and is indicative of the field’s poor understanding of the neurobiology governing CINV. Given the prevalence and clinical impact of CINV, improved understanding of the chemical mediators and pathophysiology of CINV is necessary.

Olanzapine (OLZ), historically used as an antipsychotic, was added to anti-emetic guidelines in recent years due to its efficacy in reducing both acute and delayed phases of CINV (NCCN, 2017; Navari, 2014). The anti-psychotic effects of OLZ are thought to be mediated primarily by blockade of several types of serotonin (5-HT) and dopamine receptors (Seeman, 2002; Stahl, 2013). Serotonin is well known to play a key role in emesis (Hesketh, 2008; Johnston et al., 2014) with serotonin type three receptor (5-HT3R) antagonists being among the most widely used antiemetics (NCCN; Aapro, 2005). In fact, OLZ has high affinity for 5-HT receptors (e.g., 5-HT2C, 5-HT2A, 5-HT1A, and 5-HT3 receptors; Bymaster et al., 1996; Selent et al., 2008) associated with CINV and energy balance (Clissold et al., 2013; Demirbugen Oz et al., 2018; Johnston et al., 2014; Kirk et al., 2009; Lord et al., 2017). Given that populations of neurons expressing these receptor subtypes are highly expressed in the dorsal vagal complex (DVC; Barnes & Sharp, 1999; Cabral et al., 2017; Laporte et al., 1992), a region of the hindbrain that mediates emesis in humans and other species (Babic & Browning, 2014; Hesketh, 2008; Miller & Leslie, 1994), it is possible that OLZ treatment may reduce 5-HT receptor signaling within the DVC to attenuate CINV.

Another potential mechanism by which OLZ could alleviate CINV and blunt undesirable weight loss in chemotherapy is via restoration of chemotherapy-induced declines in ghrelin secretion. Ghrelin is an orexigenic hormone mainly secreted by the stomach during fasting with high affinity for the growth hormone secretagogue receptor (GHSR; Inui, 2001; Kojima et al., 1999), which is highly expressed in central regions regulating feeding and emetic behavior (e.g., the hypothalamus (Borner et al., 2016; Zigman et al., 2006) and DVC (Cabral et al., 2017; Inui, 2001; Kojima et al., 1999). Pre-clinical and clinical evidence also demonstrate ghrelin release may reduce CINV, anorexia, and cachexia (Hiura et al., 2012; Liu et al., 2006). In this context, OLZ has been shown to increase Ghsr mRNA expression in the hypothalamus (Davey et al., 2012), to potentiate GHSR signaling in vitro (Tagami et al., 2016), and to increase circulating ghrelin levels (van der Zwaal et al., 2012; Weston-Green et al., 2011), suggesting OLZ may restore or prevent chemotherapy-induced ghrelin deficits to exert anti-emetic/anti-anorectic effects.

The present work evaluates the effectiveness of peripheral and hindbrain administration of OLZ to attenuate cisplatin-induced anorexia and pica (i.e., kaolin intake, a model for nausea/emesis (Mitchell et al., 1976; N. Takeda et al., 1993) in rats). We also examined whether OLZ administration would attenuate cisplatin-induced neuronal activation in the DVC. Finally, in an exploratory analysis to investigate potential mediators of OLZ action, we measured peripheral ghrelin levels and central mRNA expression levels of Ghsr and 5-HT receptor subunits in brain regions involved in the regulation of emesis and energy balance in rats.

Methods

Animals and Housing Conditions

Male Sprague Dawley rats (Charles River Laboratories, Wilmington, MA) were housed individually in hanging, wire-bottom cages in a 12 hr light, 12 hr dark cycle and had ad libitum access to pelleted chow (Purina Rodent Chow, 5001) and tap water except when noted. Rats were adapted to the housing conditions for 7 days before the start of the experiments. Rats in behavioral studies had ad libitum access to kaolin (Research Diets, K50001) for at least 3 days prior to the start of the experiment. All procedures conformed to the standards of the University of Pennsylvania Institutional Animal Care and Use Committee.

Drugs

Cisplatin (cis-diammineplatinum dichloride, Sigma-Aldrich) was dissolved in 0.9% saline and administered intraperitoneally (IP) in all studies at a dose of 6 mg/kg in a volume of 6 ml/kg. The cisplatin solution was sonicated and vortexed immediately before injection for each experiment. The dose of cisplatin was chosen based on previous studies showing reliable induction of kaolin intake and food intake suppression (De Jonghe et al., 2016). IP administered OLZ (OLZ, Tocris) was dissolved in 0.9% saline containing 2% acetic acid (glacial acetic acid, Sigma-Aldrich) and buffered with NaOH (pH 7) and injected at a volume of 1 ml/kg. Centrally-infused OLZ was dissolved in 100% dimethyl sulfoxide (DMSO). Doses of OLZ for all experiments were doses on in-house pilot studies showing no effect on food intake or body weight when administered alone (data not shown).

Fourth Ventricle Intracerebroventricular (Fourth ICV) Surgeries

Rats received IP anesthesia cocktail (ketamine 90 mg/kg, Butler Animal Health Supply; xylazine 2.7 mg/kg, Anased; acepromazine 0.64 mg/kg, Butler Animal Health Supply) and subcutaneous analgesia (2.0 mg/kg Metacam; Boehringer Ingelheim Vetmedica) for all surgeries. For ICV infusions, a 26-gauge guide cannula (8 mm 81C3151/Spc, Plastics One) directed at the fourth ventricle was implanted (bregma −11.6 mm, at midline, and dorsoventral −7.2 mm) and affixed to the skull with screws and dental cement as previously published (Schmidt et al., 2016). All rats were given 1 week to recover from surgery. To confirm cannula placement, blood glucose was measured following 5-thio-D-glucose (210 µg in 2 µL of artificial cerebrospinal fluid) infusion as described previously (Hayes et al., 2011; Slusser & Ritter, 1980). All rats passed verification test and were therefore included in the experiment.

Experiment 1: Effects of IP Olanzapine on Cisplatin-Induced Pica, Anorexia, and Body Weight Loss

We evaluated the ability of OLZ to ameliorate cisplatin-induced pica, anorexia, and body weight loss using a counterbalanced approach as previously described (Alhadeff & Holland, 2015; Alhadeff et al., 2017) [Figure 1(a)]. Weight-matched rats (n = 19; 380–400 g, n = 9–10/group) received IP injections of OLZ (2 mg/kg) or vehicle (2% acetic acid saline solution). In the first round, all animals received their assigned treatment (OLZ or vehicle) and 15 min later animals received IP saline (shortly before dark onset). Treatment (OLZ or vehicle) was then administered again 6, 24, and 32 hr later. In the second round (4 days after initial injection), the same injection schedule was conducted but instead animals received IP cisplatin shortly before dark onset. Food and kaolin intake measurements were taken as previously described (De Jonghe et al., 2012) at 2, 6, 24, 48, and 72 hr post IP injections (Figure 1(b), (c)). Body weights were taken daily at dark onset (Figure 1(d)).

Figure 1.

Systemic olanzapine attenuates cisplatin-induced kaolin intake (pica) and body weight loss, but not anorexia. (a) Experimental timeline used to evaluate chemotherapy-induced kaolin intake, body weight loss, and anorexia. (b) IP OLZ (2 mg/kg, twice a day for two consecutive days) decreased cisplatin-induced kaolin intake at 6, 48, and 72 hr. While IP OLZ did not attenuate cisplatin-induced anorexia (c), OLZ treatment was effective in ameliorating cisplatin-induced body weight loss at 24 hr (d). Data are expressed as mean ± SEM. * p < 0.05.

Experiment 2: Effects of fourth ICV Olanzapine on Cisplatin-Induced Pica, Anorexia, and Body Weight Loss

The paradigm used was similar to the design described in Experiment 1. Two groups of weight-matched animals (n = 4–6/group, 420–500 g) received fourth ICV infusions of OLZ (125 µg/2 µL) or vehicle (100% DMSO) followed by IP saline (shortly before dark onset). Twenty-four hours later, all rats received a second 4th ICV infusion of their assigned treatment (OLZ or vehicle). Four days later, the same rats underwent the same infusion schedule (fourth ICV infusions of OLZ or DMSO), this time followed by IP cisplatin (See Figure 2(a)). Food intake, kaolin consumption and body weight measurements were taken as in Experiment 1.

Figure 2.

Hindbrain administered (fourth ICV) olanzapine decreases cisplatin-induced kaolin intake, and weight loss, but not anorexia. (a) Experimental timeline used to evaluate chemotherapy-induced kaolin intake, body weight loss, and anorexia. ICV OLZ (125 µg) decreased cisplatin-induced kaolin intake (b) and body weight loss at 48 and 72 hr (d), but not anorexia (c). Data are expressed as mean ± SEM. *p < 0.05, **p < 0.01.

Experiment 3: Assessment of Olanzapine Treatment on Cisplatin-Induced c-FOS in DVC Neurons

Body weight-matched rats (n = 16; 250–300 g) were assigned to one of four conditions (n = 4/group) receiving either IP OLZ (2 mg/kg) or saline followed (15 min later) by IP cisplatin (6 mg/kg) or saline (prior to dark onset) in a 2 × 2, between subjects design. To avoid feeding-related changes in c-Fos expression, all animals were food deprived until perfusion 6 hr later. Brain sections containing the DVC were then processed for c-Fos as published (De Jonghe & Horn, 2009; Holland et al., 2014). Free-floating sections were washed with 0.1 M PBS (3× 8 min), incubated in 0.1 M PBS containing 0.2% Triton X-100 (PBST) and 5% normal donkey serum (NDS) for 1 hr, followed by an overnight incubation with rabbit anti-Fos antibody (s2250; Cell Signaling, 1:1,000 in PBST). After washing (3× 8 min) with 0.1 M PBS, sections were incubated with the secondary antibody donkey anti-rabbit Alexa Fluor 594 (Jackson Immuno Research Laboratories, 1:500 in 5% NDS PBST) for 2 hr at room temperature. After final washing (3× 8 min in 0.1 M PBS) the sections were mounted onto glass slides (Superfrost Plus, VWR and coverslipped with Fluorogel (Electron Microscopy Sciences). c-Fos positive neurons were visualized and quantified using fluorescence microscopy (Nikon 80i, NIS Elements AR 3.0) at 20× magnification.

Experiment 4: Assessment of Olanzapine Treatment on Cisplatin-Induced Reductions in Plasma Ghrelin Levels

Rats (n = 16; 300–350 g) received IP injections of OLZ (2 mg/kg) or saline followed 15 min later by IP cisplatin (6 mg/kg) or saline (prior to dark onset) in a between subject design (n = 4/group). Food was removed immediately before the first injection [1 hr prior to dark onset] to minimize contributions of ingestive behaviors to circulating ghrelin levels. Rats were decapitated 6 hr post-injection and trunk blood was collected into heparin-coated tubes containing protease inhibitor, p-hydroxymercuribenzoic acid (Bertin Pharma, final concentration of 1 mM). Blood was centrifuged (7,000 g for 7 min) to collect plasma and stored at −80°C until assayed. Plasma was assayed using acylated ghrelin ELISA kit and unacylated ghrelin ELISA kit (Bertin Pharma, Montigny-le-Bretonneux, France) according to the manufacturer’s instructions. Plasma samples were analyzed in duplicate.

Experiment 5: Assessment of Olanzapine Treatment on Cisplatin-Induced Changes in Central mRNA Expression of GHSR and Serotonergic Receptors (5-HT2C, 5-HT2A, 5-HT3, 5-HT1A)

Shortly before dark onset, rats (n = 32; 300–350 g) received IP OLZ (2 mg/kg) or saline followed 15 min later by IP cisplatin or saline (n = 8/group), food was removed, and animals were euthanized 6 hr later as described in experiment 4. Brains were collected and flash frozen in cold isopentane. Whole hypothalamus and bilateral micropunches (1 mm in diameter) containing regions of the DVC, parabrachial nucleus (PBN), and central amygdala (Small & DiFeliceantonio, 2019) were collected according to previously published methods (Holland et al., 2014). The collected tissue was obtained at the following coordinates: DVC (−14.3 to −13.3 mm from bregma), PBN (−10.0 to −8.5 mm from bregma), and CeA (−3.0 to −2.0 mm from bregma). RNA was extracted using Trizol (Invitrogen) and RNeasy Kit (Qiagen) according to manufacturer’s instructions. cDNA was generated using a High Capacity cDNA Reverse Transcription Kit (Applied Biosystems). Quantitative real-time polymerase chain reaction (qPCR) was performed using pre-designed Taqman probes (Applied Bioscience). Rat beta actin (Cat. 4352340E) was used as a housekeeping reference gene and probes were used to quantify relative mRNA levels of Ghsr (RN00821417), the 5-HT2C receptor gene (Htr2c; Rn00562748), the 5-HT2A receptor gene (Htr2a; Rn00567856), the 5-HT3A receptor gene (Htr3a; nRn00567878), and the 5-HT1A receptor gene (Htr1a; Rn00569876). Samples were run in duplicate and relative mRNA expression was calculated using the comparative Ct method as described previously (Holland et al., 2014).

Statistics and Data Evaluation

All data are expressed as means ± SEM. In Experiment 1 and Experiment 2, kaolin intake, food intake, and body weight were calculated by subtracting cumulative measurements at each time point from measurements taken at 0 hr. For Experiments 1 and 2, two-way ANOVAs were performed to evaluate our two drug effects (main effects: cisplatin/saline and OLZ/vehicle) on kaolin intake, food intake, and body weight change. Two-way ANOVAs for Experiment 1 and 2 were followed by a Sidak post-hoc test. For Experiments 3–5, two-way ANOVAs were conducted to examine the two drug treatments (main effects: cisplatin/saline and OLZ/vehicle) followed by Tukey’s multiple comparison test. In Experiment 3, c-Fos positive neurons were quantified. The average of three sections per animal/per plane level examined were used to quantify the total number of cFos immunofluorescent neurons. Neurons were counted for total area postrema (AP) and bilaterally for total nucleus of the solitary tract (NTS) at respective plate level. In Experiment 4, sample absorbance measurements were used to determine concentration by linear regression from standard curves. To determine relative mRNA expression in Experiment 5, average delta Ct values of beta actin were subtracted from average delta Ct values of the gene of interest. For all statistical tests, a p-value less than 0.05 was considered significant. For each experiment, statistical differences between mean values were calculated using PRISM (GraphPad Inc.).

Results

Experiment 1: Systemic Olanzapine Attenuates Cisplatin-Induced Kaolin Intake (Pica) and Body Weight Loss, but not Anorexia

There was a significant main effect on kaolin intake on cisplatin/saline IP injection at all time points measured (all Fs_{( 1,17)} ≥ 13.13; p < 0.05), but not OLZ/vehicle IP injections (all Fs = ns). Post-hoc analyses revealed a significant attenuation of cisplatin-induced kaolin intake by OLZ vs. vehicle at 6, 48, and 72 hr (Figure 1(b), all ps < 0.05). We also observed significant interaction between cisplatin/saline and OLZ/vehicle IP injections on body weight at 24 hr (F_{( 1,17)} > 5.07; p < 0.05), but not at 48 or 72 hr [all Fs = ns]. Post-hoc analyses revealed an attenuation of cisplatin-induced body weight loss in animals pre-treated with OLZ vs. vehicle at 24 hr (Figure 1(d), p < 0.05). While we did see a main effect of cisplatin/saline injection on food intake at 6, 24, 48, and 72 hr time points (all Fs F_{( 1,17)} ≥ 17.29; p < 0.0001), no discernable effects of OLZ were revealed on food intake (Figure 1(c)).

Experiment 2: Hindbrain Administered (Fourth ICV) Olanzapine Decreases Cisplatin-Induced Kaolin Intake, and Weight Loss, but not Anorexia

Overall, we observed a similar pattern of results in Experiment 2 compared to experiment 1, with cisplatin/IP injection main effect on kaolin intake at each time point measured (all F_{( 1,17)} ≥ 66.44; p < 0.05) and significant interaction effect noted for IP × ICV drug treatments on body weight at 48 and 72 hr (both F_{( 1,17)} ≥ 5.33; p < 0.05). Post-hoc analyses showed significant decreases in cisplatin-induced kaolin intake and body weight loss in animals pre-treated with OLZ vs. vehicle at 48 and 72 hr (Figure 2(b), p < 0.05). There were no significant changes to cisplatin-induced anorexia following either ICV treatment [Figure 2(c), All Fs = ns].

Experiment 3: Systemic Olanzapine Decreases Cisplatin-Induced c-FOS in NTS, but Not AP, Hindbrain Neurons

Representative immunofluorescent sections from the DVC 6 hr post cisplatin or saline IP administration are depicted in Figure 3(a). Generally, cisplatin vs. saline IP treatment induced much greater c-Fos expression in the NTS at all 3 plate levels (−14.3, −13.8, −13.3 mm, Figure 3(a), (b)). There was a significant IP × IP treatment interaction on c-Fos positive bilateral counts of neurons in the NTS at −13.8 and −13.3 mm (all F_{( 1,12)} ≥ 8.33; p < 0.05) with post hoc tests revealing that cisplatin-induced c-Fos immunofluorescence in the NTS was significantly attenuated by OLZ vs. saline pre-treatment (Figure 3(a), (b) p < 0.05). We did see a cisplatin/saline main effect on c-Fos expression in the AP (F_{( 1,12)} ≥ 5.18; p < 0.05) however, no olanzapine/saline injection effect was observed in the AP [F = ns].

Figure 3.

Systemic olanzapine decreases cisplatin-induced c-Fos in NTS, but not AP, hindbrain neurons. (a) Representative immunofluorescent sections of the DVC showing c-Fos positive cells 6 hr after treatment across different plane levels (distance from Bregma). (b) Cisplatin (6 mg/kg) induced c-Fos immunofluorescence in DVC neurons 6 hr following administration. IP OLZ pre-treatment (2 mg/kg) reduced c-Fos immunofluorescence in the NTS (−13.8 mm, −14.3 mm) in cisplatin-treated rats (OLZ-CIS vs SAL-CIS). In the AP, c-Fos immunofluorescence was similar among cisplatin-treated animals, regardless of OLZ or vehicle pretreatment. Data are expressed as mean ± SEM. Different letters within the same panel represent significant differences between lettered groups (p < 0.05).

Experiment 4: Systemic Olanzapine Prevents Cisplatin-Induced Reductions in Circulating Ghrelin Levels

There was a significant IP × IP injection interaction on acylated (active) ghrelin levels (F_{( 1,12)} = 10.63; p < 0.01) and acylated to unacylated ghrelin ratio (F_{( 1,12)} = 11.09; p < 0.01) 6 hr post treatments. Post hoc tests showed cisplatin-induced declines in acylated ghrelin and acylated to unacylated ghrelin ratio was blocked in animals pre-treated with OLZ (Figure 4(a), p < 0.05). No changes in unacylated ghrelin were observed in any treatment condition (p = ns).

Figure 4.

Systemic olanzapine prevents cisplatin-induced reductions in circulating ghrelin levels. (a) Cisplatin treatment (6 mg/kg IP) reduced circulating acylated ghrelin levels 6 hr after administration. OLZ (2 mg/kg IP) blocked cisplatin-induced decreases in circulating acylated ghrelin. (b) There were no significant differences in unacylated ghrelin levels between treatment groups. (c) IP OLZ blocked cisplatin-induced decreases in the ratio of acylated to unacylated ghrelin. Data are expressed as mean ± SEM. Different letters within the same panel represent significant differences between lettered groups (p < 0.05).

Experiment 5: Systemic Olanzapine Blocks Cisplatin-Induced Increases of DVC and Hypothalamic Htr2c Expression

Our results show a significant IP × IP treatment interaction on 5-HT2C receptor gene, Htr2c, in the DVC (F_{( 1,23)} > 44.45; p < 0.05) and hypothalamus (F_{( 1,25)} > 5.09; p < 0.05) where post hoc tests showed that OLZ pre-treatment prevented cisplatin-induced increases in Htr2c expression in both regions (Figure 5(a), (b), all ps < 0.05). There were no significant differences in Htr2a, Htr1a, nor Htr3 mRNA levels in the DVC (Figure 5(b)), hypothalamic Htr2a expression (Figure 5(b)), nor Htr2c and Ghsr expression in the PBN and CeA, across all treatments [all ps = ns].

Figure 5.

Systemic olanzapine blocks cisplatin-induced increases of DVC and hypothalamic Htr2c expression. OLZ (2 mg/kg IP) blocked cisplatin-induced increases in the expression of Htr2c in the DVC (a) and hypothalamus (b) 6 hr post IP cisplatin (6 mg/kg IP) or saline treatment. There were no significant differences in Htr2a, Htr1a, and Htr3 mRNA levels in the DVC (a), hypothalamic Htr2a expression (b), nor Htr2c and Ghsr expression in the CeA (c) and PBN (d) across all treatments. Data are expressed as mean ± SEM. Different letters within the same panel represent significant differences between lettered groups (p < 0.05).

Discussion

The current set of studies examined the potential positive effects of OLZ on cisplatin-induced sickness behavior and energy balance dysregulation in rats. Behavioral experiments showed that prophylactic IP or fourth ICV OLZ reduces cisplatin-induced pica and partially ameliorates body weight loss following cisplatin chemotherapy treatment. Our immunofluorescence data suggest that these behavioral improvements may be partially mediated by reduced activation of NTS, but not AP, neurons within the hindbrain. Plasma measurements of ghrelin also suggest that IP OLZ restores normative levels of circulating active ghrelin that are reduced by cisplatin treatment. Lastly, we showed in exploratory studies that cisplatin-induced elevations in central mRNA expression of Htr2c in the hindbrain and hypothalamus are blocked by systemic OLZ administration.

The temporal profiling in the reduction in cisplatin-induced pica following IP and fourth ICV OLZ suggests that systemic OLZ may be effective at reducing sickness behavior in acute (6 hr) and delayed phases (48–72 hr) of CINV, however, the central sites of action contributing to OLZ’s anti-emetic side effects are not fully defined. Physiological data have demonstrated that OLZ decreases the excitability of neurons located in the DVC, and reduces vagal output to subdiaphragmatic organs (Anwar et al., 2016). It has long been known that cisplatin increases neuronal activation in the DVC, an effect that is dependent on vagal input and contributes to nausea and emesis (Horn, 2009). To this end, our fourth ICV OLZ administration data showing that direct hindbrain delivery of OLZ attenuates cisplatin-induced pica, in conjunction with our immunofluorescence data demonstrating that IP OLZ produces a decline in NTS c-Fos (in cisplatin-treated rats compared to vehicle injected controls), suggests that the hindbrain is at least a partial mediator of OLZ’s anti-emetic action in decreasing the severity of sickness behavior.

More generally, these data also support the use of rats as a model to address gaps in the literature examining the efficacy of olanzapine and other anti-emetics. For example, only one previous study by Machida et al. (2015) examined systemic OLZ on malaise in young Wistar rats and observed only a marginally significant cisplatin-induced pica, making determinations of OLZ effectiveness limited. We report IP OLZ significantly reduced cisplatin-induced pica (∼30% reduction) at 24 hr and both IP and ICV OLZ decreased pica (by ∼28% and ∼42%; respectively) 48 hr following cisplatin treatment, suggesting that OLZ successfully reduces pica under more robust experimental conditions (i.e., when a strong and reliable induction of cisplatin-induced pica is observed).

When comparing literature with other known anti-emetics, including the classical 5-HT3 receptor antagonists (5-HT3RAs), we find similar magnitude reductions in cisplatin-induced pica as reported here due to OLZ pretreatment. Jeong et al. (2005) found 5-HT3RA tropisetron (10 mg/patch), reduced pica by ∼42%, 48 hr following cisplatin (10 mg/kg IV) treatment, a reduction consistent with our observations in ICV OLZ treated rats under similar conditions. However, the effectiveness of other 5-HTRAs and Neurokinin-1 receptor antagonists (NK-1RAs) on acute (0–24 hr) cisplatin-induced pica is unclear, as effectiveness, drug dosing, and route of administration vary across the literature. Jeong et al. (2005) and Yamamoto et al. (2014) found a 56.6% and a 74% reduction, respectively, in acute (0–24 hr) cisplatin-induced pica when rats were pretreated with 5-HT3RAs tropisetron (10 mg/kg patch) and granisetron (0.05 and 0.1 mg/kg IP, four times a day). In contrast, others (Rudd et al., 2002) found 5-HT3RA, ondansetron (2 mg/kg IP), increased cisplatin-induced acute (0–24 ho) pica (41% increase compared to vehicle-cisplatin treated rats). The efficacy of NK-1RAs in reducing cisplatin-induced pica also varies across studies. Yamimoto et al. (2014) also found an almost 100% reduction in acute and delayed cisplatin-induced pica when rats were pretreated by NK-1RA, aprepitant (1 and 2 mg/kg IG). However, Goineau and Castagne (2016) found aprepitant (10 and 30 mg/kg PO) did not reduce cisplatin (6 mg/kg IP) induced pica in rats at any time point. Further studies clarifying the role of guideline-recommended CINV antiemetics in preclinical models would be useful when using standardized dosing and combination therapies (including olanzapine, 5-HTRAs and NK-1RAs alone and in combination).

We did not see a significant reduction in cisplatin-induced anorexia in animals pretreated with OLZ. This is not entirely surprising as the doses of OLZ used in our studies were chosen due to the subthreshold nature of these doses on energy balance under non-pathological conditions. Previous work using acute and chronic OLZ administrations, at doses comparable to what we used in our current study, failed to significantly affect food intake in male rodents (Cooper et al., 2007). Cisplatin-induced weight loss was attenuated following both IP and ICV OLZ. A reduction in weight loss without significant changes in food intake may be due to metabolic changes (e.g., decreased energy expenditure/thermogenesis) independent of caloric intake. Although we did not measure energy expenditure, our findings are nonetheless consistent with work showing olanzapine-induced metabolic changes, including weight gain, hypothermia, and decreased locomotor activity, in the absence of hyperphagia (Evers et al., 2010). A possible feeding-independent adipogenic action of OLZ could be due to an up-regulation of lipogenic enzymes and a reduction of lipid export in adipose tissue (Albaugh et al., 2011). Another factor attenuating wasting/body weight loss may be a reduction of metabolic rate, which is a well-documented effect of serotoninergic-based anti-psychotic drugs (Albaugh et al., 2011).

To further explore likely chemical pathways impacted by OLZ in our model, we focused on ghrelin and serotonin as potential biological mediators of the positive effects observed by OLZ treatment in cisplatin-treated animals. The positive effects of ghrelin or its analogs administration in the prevention of CINV, anorexia, and cachexia have been well documented (Hiura et al., 2012; Liu et al., 2006; Rudd et al., 2006). However, ghrelin-based clinical trials have largely failed to demonstrate significant benefits of ghrelin on the aforementioned pathophysiological processes (Currow et al., 2017; Temel et al., 2016). This discrepancy may be caused by partial ghrelin resistance, which may limit the effectiveness of ghrelin-based therapy in the treatment of cancer cachexia (Terawaki et al., 2017). Chronic administration of OLZ (at oral doses ranging from 2.5–20 mg) has been reported to increase ghrelin levels in psychiatric patients (Murashita et al., 2005; Zhang et al., 2013), highlighting the possibility that OLZ might counteract ghrelin insensitivity. Here, our data show that IP OLZ reverses cisplatin-induced reductions in circulating ghrelin, presumably beneficially impacting gastrointestinal function among other potential positive physiological effects. We did not, however, observe Ghsr mRNA expression alterations following OLZ treatment in any brain region tested. Nevertheless, OLZ may be able to counteract ghrelin-resistance, highlighting the possibility of combination therapy of OLZ and ghrelin may be more effective at managing CINV (Rudd et al., 2018).

We also found that IP OLZ reversed cisplatin-induced increases in Htr2c gene expression in the DVC and hypothalamus. Cisplatin-induced increases in Htr2c expression in these regions have been associated with nausea and anorexia (Kirk et al., 2009; Silenieks, 2014; H. Takeda et al., 2008). Conversely, OLZ is known to cause weight gain via 5-HT2C receptor antagonism (Huang et al., 2018; Lord et al., 2017), and inhibition of 5-HT2C receptors has been linked to decreases in nausea (Okada et al., 1995; H. Takeda et al., 2008). These studies are consistent with our findings and suggest that OLZ’s antiemetic effect may be mediated in part by attenuating cisplatin-induced effects on 5-HT2C signaling. Further studies are necessary to directly examine whether OLZ’s anti-emetic effects are dependent on 5-HT2C signaling and whether 5-HT2C antagonism could help manage CINV. Intriguingly, 5-HT2C receptor activation is known to regulate ghrelin as part of a negative feedback loop that modulates energy balance (Nonogaki et al., 2006; Schellekens et al., 2013, 2015). Thus, future work examining whether 5-HT2C and ghrelin can independently modulate cisplatin-induced side effects would be impactful to resolve this question.

In summary, we found that OLZ reduces sickness behavior and weight loss following cisplatin treatment in rats. We also found that OLZ decreased cisplatin-induced neuronal activation in the NTS, rescued cisplatin-induced deficits in circulating ghrelin, and reversed cisplatin-induced increases in Htr2c expression in the DVC and hypothalamus. Taken together, our findings suggest that serotonergic (5-HT2C) and ghrelin systems may be possible mediators of OLZ’s anti-emetic properties, evidence which may enhance development of more effective and better tolerated anti-emetic pharmacotherapies focusing on these biochemical pathways.

Nursing Implications

There is a need to better understand the mechanisms underlying symptoms and the important role that nursing science can play in achieving this mission. While the National Institute of Nursing Research supports the advancement of symptom science, CINV—and nausea as a standalone symptom—remain neglected in the context of care and research. Despite advancements in the treatment and management of CINV, nausea and vomiting continue to be cited among the most distressing, intolerable, and feared side effects of chemotherapy (Aapro, 2018; Ilyas et al., 2020; Ruhlmann et al., 2015). Thus, it is imperative for nurses, clinicians, and scientists to better understand the biological mechanisms underlying CINV. Identifying the mediators underlying the anti-emetic properties of OLZ can help identify more effective and better tolerated treatments that can ultimately increase the quality of life of people undergoing chemotherapy.