Introduction

Food intake and weight are important indicators of an animal’s overall physical and psychological health.⁴³ Animals in biomedical research may occasionally develop hyporexia (decreased appetite) or anorexia (complete loss of appetite). Hyporexia is a clinical sign secondary to a primary cause, which can be multifactorial. Animals in biomedical research are used as models for various disease conditions that are often achieved by induction or selective breeding. Animals can also develop spontaneous disease. Conditions such as cancer and viral infection can also reduce appetite and body weight.^4,6 Inflammatory cytokines are a possible mechanism for anorexia associated with illness. These cytokines interact with the central nervous system and can disrupt the hunger pathway of the hypothalmic-pituitary axis.^4,6 Stress can also disturb both the hypothalamic-pituitary-adrenal (HPA) axis function and feeding behavior.^48,72 Animals in biomedical research are exposed to a variety of stressors, including experimental manipulation,³¹ navigating a social hierarchy,^29,65,69 or living with a spontaneous or induced disease.^35,38,41 Certain models require surgery as part of study design. These animals often develop hyporexia or anorexia after surgery. Some sedatives and anesthetics, including ketamine and isoflurane, can also reduce appetite after surgery.^5,60,71 In addition, some analgesics like buprenorphine may reduce food intakes.^27,67 Various methods are used to mitigate and alleviate pain in research animals,⁴⁷ but in some cases, animals may experience breakthrough pain. Pain is another factor that can reduce food intake.^8,42 Long episodes or frequent episodes of inappetence or hyporexia can lead to weight loss, which can affect both animal welfare and research. Many institutions consider a reduction in body weight of 20 to 25% or more as a criterion for humane endpoint consideration.⁶²

Treatment of inappetence is important because poor nutritional intake can worsen disease-related catabolism,⁵⁶ leading to nutritional deficiencies,^13,25,30 altering research^10,46 and promoting weight loss. Treatment of inappetence can be challenging and unsuccessful. Treatment is usually centered around resolving any underlying disease process, encouraging increased voluntary food intake by providing highly palatable or valuable food items, or pharmacological treatment to stimulate appetite. Additional options include providing enteral nutrition via oral or nasogastric tubes;⁶⁴ however, doing this may require sedation or anesthesia. If loss of appetite is a known factor in specific models, surgical placement of an indwelling feeding tube can be attempted.^14,33 Unfortunately, efficacy and long-term success can be variable with these methods. Interventions that do not require sedation are preferable because sedation alone can reduce food intake.^46,56 Determination of the primary cause of hyporexia can allow this problem to be addressed before significant weight loss occurs. This might be best achieved by using effective appetite-stimulating drugs. Several medications are known to have either direct or indirect effects on increasing food intake, but efficacy varies by mechanism of action and by species.

Several medications are currently FDA-approved for the treatment of hyporexia in domestic species,^45,51,75 while other agents are used off-label and lack formal approval.¹ Reliance on safety and efficacy data that are extrapolated from domestic species means that many therapies are used off-label for treatment of diverse species in veterinary medicine.⁵⁴ Some medications have been used to treat inappetence in NHPs. Benzodiazepines such as diazepam and midazolam increase appetite in rhesus macaques, baboons, and marmosets.^20,26 Benzodiazepines inhibit the neurotransmitter GABA and affect appetite by improving palatability.^12,22 Megestrol acetate (MA) is another drug that has been used as an appetite stimulant in NHPs. MA is a synthetic progestin. In 1994 the FDA approved MA for treating anorexia, cachexia, and/or weight loss in human patients diagnosed with AIDS.⁷³ The exact mechanism by which MA promotes weight gain is not clear, although several mechanisms have been proposed that are related to its action on progesterone and glucocorticoid receptors.⁷³ Tyrosine protein kinase receptor agonists have also showed promising effects on appetite stimulation in several NHP species when administered by subcutaneous injection, but their availability is limited for animal use.³⁷ Cyproheptadine, a serotonin 5-HT2 and histamine H1 antagonist used for the treatment of allergies, is another medication which has been used as an appetite stimulant.²⁴ Mirtazapine is an antidepressant which has similar effects to cyproheptadine, antagonism of the serotonin 5-HT2 receptor.⁴⁶ Serotonin interacts within the hypothalamus with endogenous orexigenic (NeuropeptideY/Agouti related protein) and anorectic (α-melanocyte stimulating hormone) peptides,⁶⁶ which through a cascade of processes ultimately result in satiation.²³ Anecdotally, all these therapeutics have been used by various facilities to treat inappetence in NHPs, with varying levels of success.

Capromorelin acts by a completely different mechanism than these other appetite stimulants. It is the only Food and Drug Administration (FDA) approved appetite stimulant for use in dogs and cats.^53,76 Capromorelin is an orally active growth hormone secretagogue receptor (GHS-R) agonist that mimics the effects of endogenous ghrelin, a 28 amino acid peptide hormone, which is primarily synthesized in the stomach.¹⁷ Ghrelin has a relatively short half-life.⁷⁶ Plasma levels of ghrelin spike in anticipation of eating and fall rapidly after eating. Ghrelin, its receptor, and its gene are relatively well conserved across species.^45,57,58 When GHSR-agonists such as capromorelin bind to receptors in the arcuate nucleus of the hypothalamus, they initiate a cascade of neuroendocrine mechanisms that stimulates appetite,^53,75 resulting in an increase in food consumption⁵⁹ and a gain in body weight.

The mammalian hunger pathway is a complex of neuronal and endocrine interactions. At its center lies the arcuate nucleus of the hypothalamus.³ Within this nucleus, peripheral peptides such as ghrelin cross the blood brain barrier and interact with neurons, stimulating the release of other peptides related to food intake and weight gain. For this reason, ghrelin has been described as the hunger hormone.^15,16 Studies have shown that ghrelin increases growth hormone (GH) secretion, which in turn increases serum levels of insulin growth factor 1 (IGF-1) from the liver.^7,49 GH is the primary factor affecting circulating IGF-1 levels,³² but other hormones such as leptin, ghrelin and insulin can modulate the physiologic effects of IGF-1 on growth.⁴⁰ Certain disease states such as hypoproteinemia and diabetes mellitus can also influence IGF-1 levels.^21,36 GH release is regulated in part by means of a negative feedback loop with IGF-1.¹¹ IGF-1 is a peptide hormone that is essential to the prenatal and postnatal growth periods in humans and many mammalian species.⁵⁵ IGF-1 is involved in metabolism, tissue regeneration and disease pathogenesis.⁵⁵

Capromorelin does not increase circulating ghrelin or enhance the effects of ghrelin; instead capromorelin acts as a ghrelin receptor agonist. In this respect, capromorelin is more akin to synthetic ghrelin. Administration of ghrelin receptor agonists like capromorelin have been shown to cause appreciable weight gain in dogs, cats, chickens, rabbits, and rodents.^{9,63,70,75,76} When administered orally, capromorelin is absorbed from the gastrointestinal tract into the blood stream, similar to endogenous ghrelin. In this respect, it is analogous to the natural physiology of ghrelin.⁷⁶ While ghrelin is relatively short acting, capromorelin has a longer half-life and therefore, more sustained effects.⁷⁵ Aside from increases in weight, measurements of growth hormone and IGF-1 can be used to evaluate the physiologic response to capromorelin administration because these 2 hormones act downstream of the GHS-R.⁷⁵ Growth hormone secretion is pulsatile and fluctuates throughout the day, so it is difficult to measure accurately as compared with IGF-1, whose levels do not fluctuate as frequently.²

While currently FDA-approved only for use in dogs and cats, anecdotal reports from clinical veterinarians at the Emory National Primate Research Center (ENPRC) suggest that capromorelin may be an effective appetite stimulant in rhesus macaques. It has been successfully used at ENPRC to increase weight in animals experiencing weight stagnation or weight loss. The use of capromorelin in NHP is considered extralabel and no published reports have investigated its efficacy or safety for use in NHP. In dogs, reported side effects include diarrhea, vomiting, hypersalivation, hyperactivity and polydipsia. Increases in blood urea nitrogen, phosphorus and creatinine have also been reported.^19,74

The capromorelin dose used at ENPRC is extrapolated from canine studies.^75,76 In this study we tested the hypothesis that capromorelin (ENTYCE, Elanco US, Greenfield, IN) administered orally to rhesus macaques once daily at 3 mg/kg for one week will result in a statistically significant increase in body weight and IGF-1 levels as compared with untreated macaques. We also hypothesized that this short-term administration of capromorelin would not result in any significant changes in blood parameters indicative of major organ dysfunction or hematopoietic disruption. We achieved this by evaluating blood before and after drug administration to the macaques.

Materials and Methods

Study design.

The study took place over 14 d, with days 1-7 being a baseline period and days 8 to 14 being the actual study period (Figure 1). Macaques were fed slight modifications to the ENPRC “Feeding of Nonhuman Primates” SOP. Animals were still fed twice daily, with the bulk (75 to 80%) administered in the morning, but each animal received approximately 5 to 10 more biscuits than they would regularly receive in accordance with the aforementioned SOP in anticipation of an increased appetite. This resulted in a total of 20 biscuits a day for females and 30 biscuits a day for males. Each morning, the remaining biscuits in the cage were removed and counted before new chow was provided. For ease of counting, only quarter biscuits, half biscuits and full biscuits were counted. Crumbs that were smaller than a quarter of a biscuit were discarded. On days when animals were sedated, they received their full ration of biscuits (morning and afternoon) and dietary enrichment in the afternoon after they had fully recovered from anesthesia. Animals received a physical exam on every day of sedation. Chow was pulled 2 to 4 h prior to sedation. Animals had free access to water through an automatic watering system throughout the duration of the study.

View Figure

• Download as Powerpoint
• Download Figure

On day 8, all macaques were sedated for oro-gastric administration of Capromorelin and serial and blood collections, the details of which can be found below. Following recovery from the final sedation event, animals were provided with their total allotment of food for that day for the final 6 d of the study, macaques were fed essentially as much the same as they had received during the first 7 d of the study, except that treated animals received 3 mg/kg of capromorelin in a highly palatable food item once daily, before receiving their normal morning rations. Macaques were observed 15 to 30 min after being offered these palatable, treated food items. If items were not consumed within this time frame, animals were moved to the front of the cage by using a squeeze cage mechanism and their daily capromorelin dose was administered directly into the mouth by using a gavage needle. Control animals received palatable food items laced with a volume of water equivalent to what their dose of capromorelin would have been. On day 11 and 14 of the study, animals were sedated 4 h after capromorelin administration for further blood collection. On these two days, animals received their total ration of food once recovered.

Results

Figure 1 suggests that the mean weights by study group were similar over time. The observed difference in baseline weight reflected in Figure 1 suggests a possible need to adjust for baseline weight. Figure 2 shows mean weights of study animals and indicates the importance of adjusting for baseline weight. The 95% confidence intervals (CIs) for the two study groups did not overlap on days 11 and 14, suggesting differences that may be both statistically and biologically important. Additional appropriately powered studies are needed.

View Figure

• Download as Powerpoint
• Download Figure

Figure 3 shows that after adjustment for baseline weight, body weights changed significantly but in different ways during the 14-d follow-up (P < 0.001 for the interaction between study group and time). Mean adjusted weights were not significantly different between the treated and control groups at baseline (10.0 kg and 10.8 kg for the capromorelin and control group respectively; P = 0.59). Weight was significantly higher in the treatment group on days 11 (10.4 kg and 10.1 kg for treated and control groups, respectively; P = 0.006) and 14 (10.5 kg and 10.1 kg for treated and control groups, respectively; P = 0.004). The adjusted P values on days 11 and 14 were 0.06 and 0.05, respectively. The mean differences between groups (treated minus control) were 0.33 kg (95% CI, 0.10 to 0.55) on day 11 and 0.38 kg (95% CI, 0.13 to 0.62) on day 14. The different temporal patterns suggest that the change over time was different between the two study groups. Both Figures 2 and 4 show the temporal patterns for each group.

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

IGF-1 in the control and treated groups changed in significantly but in different ways during the 7 d of follow-up (P = 0.01 for the interaction of group and time) (Figure 4). Mean IGF-1 levels were not significantly different between the groups at baseline (775 ng/mL and 549 ng/mL for the treated and control groups, respectively; P = 0.08) but was significantly higher in the treated group on days 11 (859 ng/mL and 516 ng/mL, P = 0.007) and 14 (842 ng/mL and 513 ng/mL, P = 0.009).

After adjusting for baseline IGF-1, IGF-1 was significantly higher in the capromorelin group on day 11 (775 ng/mL and 642 ng/mL, respectively, for the treated and control groups, P < 0.001) but was not significantly different on day 14 (757 ng/mL and 640 ng/mL, respectively, for the treated and control groups) (Figure 5. Mean differences (treated minus control) on days 11 and 14 were 133 ng/mL (95% CI, 50 to 216) and 118 ng/mL (95% CI, -40 to 275), respectively.

View Figure

• Download as Powerpoint
• Download Figure

An effect of sex was also statistically discernible for IGF-1 based on the adjusted analyses. IGF-1 levels were higher in males than in females (mean = 794 ng/mL for males and 535 ng/mL or females, P = 0.015)

Tables 1 to 5 show the longitudinal variability in weight and IGF-1 for the study. The tables provide the relative variability between animals and between longitudinal measures on the same animal (that is, the within-animal standard deviation). These estimates of standard deviation will be useful for power analysis for future studies because variability estimates are essential for sample size calculations. High within-animal standard deviations (relative to between-animal standard deviations) indicate that repeated measurements on the same animal will increase the power of any longitudinal study. Low within-animal standard deviations suggest that a better study design would be to use more animals and reduce the number of repeated measurements on each animal.

Table 1 shows the adjusted P values by the Tukey-Kramer test for multiple comparisons. The test compares the difference between each pair of means with appropriate adjustment for multiple testing. The pairwise comparisons are based on the studentized range, sometimes called the “honestly significant difference” test.^44,68 The time-averaged mean is the mean pooled across the five time points estimated by the mixed model analysis of repeated measures. The time-averaged IGF-1 means were 798 ng/mL and 531 ng/mL for trated and control animals, respectively. The time-averaged mean difference in IGF-1 between treated and control animals was 267 ng/mL with a standard error of the mean of 118 ng/mL. The model-based time-averaged IGF-1 mean difference is the overall effect of study group. The values for comparisons of control and treated groups are as follows: P = 0.68 for time-averaged mean; P < 0.001 for time; P < 0.001 for the group by time interaction (Table 2). For the baseline adjusted mean and 95% CI for weight (kg) by study group and days on study P = 0.06 for time-averaged mean (Control compared with Experimental), P < 0.001 for time effect, P < 0.001 for the group by time interaction and P < 0.001 for baseline adjustment (Table 3). For the mean and 95% CI for IGF-1 (ng/mL) by study group and days on study P = 0.05 for time-averaged mean (531 Control compared with 798 Experimental) P = 0.23 for time effect P= 0.003 for the group by time interaction (Table 4). For baseline adjusted mean and 95% CI for IGF-1 (ng/mL) by study group and days on study P = 0.01 for time-averaged mean (652 for Control compared with 720 for Experimental), P = 0.15 for Time effect, P = 0.001 for the group by time interaction, P < 0.001 for baseline adjustment (Table 5).

Chow counts were performed on all animals for all 14 d of the experiment (Table 6). Days before and after treatment were observationally compared to estimate changes biscuit consumption. Among the treated monkeys (1 to 6), only #2 did not consume more chow during the second week. Monkey #1 had higher chow consumption on 5 of the 7 d during the second week, and monkeys #3, #4, #5 and #6 each had 3 d on which they consumed more chow during the second week. By comparison, control monkeys 7, 8, 9, and 10 respectively had 1, 2, 3, and 4 days of greater chow consumption during the second week.

Discussion

Based on the weights of the 6 treated animals, capromorelin appears to be a promising pharmacological agent for increasing weight in rhesus macaques. Over 7 d of treatment, weight was significantly higher in the treated group as compared with the control group on days 11 and 14 (Figure 3). The weight gain observed in our treated cohort seems insignificant (0.38 kg), but it perhaps reflects the short duration of the treatment phase in combination with sedation on 3 out of 7 d treatment phase. Ketamine sedation has been linked to decreased appetite⁵⁰ and could have blunted the degree of weight gain. By comparison, macaques in the control group were also sedated at the same intervals and lost weight. On days when animals were sedated (1, 4, 8, 11, and 14), they showed a marked decline in appetite. This was most apparent on day 8, when animals were sedated 3 separate times. Three of the 10 animals consumed no biscuits’ 2 of the 3 were control animals. Another 4 animals consumed fewer than 5 biscuits on day 8. Similar reductions in chow consumption occurred on days 1, 4, 11, and 14. In this study repeated ketamine sedation was necessary to perform blood collection at the designated time points. Two of the 6 treated macaques showed an increase in chow consumption between the 2 wk, whereas the other 4 did not. One of the control animals showed increased consumptions between the 2 wk. However, regardless of chow intake, the data showed a significant weight gain in the treated as compared with the control group.

Several factors could account for the disparity between chow counts and weight. First, intake of other food items (dietary enrichment and produce) was not measured due to the difficulty in quantifying amounts consumed. This is a limitation of this study. However, all animals received the same type of enrichment and produce on a day-by-day basis. Another factor that may have contributed to the divergence of chow counts and weights is the method used to perform chow counts. Biscuits were quantified by quarter, half, and full pieces. Anything smaller than a quarter biscuit was not counted due to difficulty in accurately equating these pieces to pieces of a complete biscuit. These smaller pieces may have added up to multiple biscuits, perhaps showing an actual increase in chow consumption in either group.

The safety profile of acute capromorelin administration was assessed through a combination of blood analysis, daily observations and 5 physical exams throughout the course of the study. Blood was analyzed only for treated animals. Control animals appeared to be healthy. No signs of organ dysfunction (for example, jaundice) were noted prior or during the study. None of the macaques showed rapid changes in weight. Stool and urine output observationally appeared to be normal. Control animals had blood analysis and physical exams performed within 6 mo before the start of this study. None of their previous blood analysis indicated organ dysfunction or endocrinopathies. Many of the reported side effects of capromorelin are gastrointestinal in nature.^19,74 Over the duration of this study, none of the animals showed evidence of emesis, diarrhea, oliguria, polyuria or hyperexcitability. The only adverse effect observed in our study was hypersalivation, which has also been reported in dogs.^19,74 This hypersalivation was transient, lasting no more than a minute or 2, and was only seen in animals that were dosed directly into the mouth by using a syringe. Hypersalivation was never observed when the capromorelin was consumed with food. All 5 physical exams performed on all animals were within normal limits, with no significant findings. Blood analysis revealed multiple instances in which values were outside of the accepted reference ranges. In many of these cases, the values fell just outside the reference range. In other cases, parameters were above or below those measured at baseline, with more normal values measured in blood collected at end of study. All in all, we found no instances in which animals appeared to be experiencing an organ dysfunction or endocrinopathy, either at baseline or at the end of the study. Blood urea nitrogen, phosphorus and creatinine were reported to increase in some capromorelin treated dogs.^19,74 None of those values were increased in any of our animals. Based on all these finding, capromorelin appears well tolerated by rhesus macaques, at least as administered.

The gene that codes for IGF-1 is relatively well conserved among vertebrates but is even more closely conserved among primates. A comparison of different primate species, including humans, found a high degree of conservation of nucleotide sequences and promoter regions.⁵⁵ In rhesus macaques, IGF-1 levels follow a similar developmental pattern as in humans. Serum IGF-1 levels are low during infancy, peak during puberty, and then appear to plateau after 4.5 to 5 y of age.^39,61 Along with age, serum levels of IGF-1 appear to be lower in rhesus females than in males, with the exception that levels are higher in pregnant females.⁶¹ These factors were considered when selecting animals for this study. None of the females in this study were pregnant, and all animals but one animal were over 5 y of age, except for 1animal that was 4.5 y old. We have found that the prime rapid growth phase for rhesus macaques is between 3 to 5 y of age. Growth continues beyond 5 y of age but slows and plateaus around 7 y of age.⁵² To limit extraneous variables in our study, we chose not to use animals that had already been assigned to endocrinology or gastrointestinal research studies. Given the shortage of NHP for biomedical research, this meant that we had a limited number of animals available for our use. During the selection process, we were careful to select animals within certain age and weight ranges in order to avoid introducing confounding variables.

In addition to weight gain, IGF-1 levels were another objective measure of capromorelin efficacy. Using time points similar to those used in a canine study,⁷⁵ we found differences in IGF-1 levels, comparatively. In the canine study,⁷⁵ IGF-1 levels rose for up to 8 h after administration before returning to baseline. In our study, 4 of the 6 treated macaques had IGF-1 baseline levels at 4 h after administration; the other 2 had fallen below baseline by 8 h after administration. However, in our study, IGF-1 levels attained those reported in canine study. This difference in outcomes could be due to metabolic differences between macaques and dogs, or perhaps macaques require a higher dose than do dogs. We did not perform a comparative pharmacokinetic study, and so we do not know whether a 3 mg/kg dose results in equivalent serum levels of IGF-1 in macaques and dogs. However, as in similar to studies in dogs, cats, humans, and other species,^63,70,75,76 capromorelin administration resulted in weight gain in rhesus macaques, which was likely related to IGF-1 levels.

Capromorelin has been used at the ENPRC since 2019. It is mainly administered to animals used in transplant and infectious disease studies. Animals experience frequent sedation and experimental intervention on these studies, and inappetence and weight loss occur frequently. Other appetite stimulants used at ENPRC include megestrol acetate, cyproheptadine and mirtazapine. Frequently capromorelin was successful after these drugs had failed. Many of the animals that received capromorelin at ENPRC were treated with it for weeks to months at a time. During these treatment periods, blood analysis was frequently performed as part of their study protocols. Just as we saw in this acute study, long term treatment did not appear to be associated with variations in blood parameters. As part of this study, retrospective weight data was collected from 43 animals treated with capromorelin in 2019. This data was not included in our report because several variables negated our ability to statistically analyze the data. First, these animals typically had previously received various appetite stimulants prior to being started on capromorelin. In most cases, those other appetite stimulants were continued along with capromorelin administration. Secondly these animals were sedated at different time points from one another, and when being used on a protocol they underwent a variety of invasive procedures, any of which could have affected food intake. Third, we had no data from control animals. These confounding variables prevented it from contributing to confident conclusions.

Another limitation influencing the data collected is that the carpromorelin is not palatable to rhesus macaques. We attempted to mask the taste with peanut butter, Nutella, fruit flavored yogurt and SyrPalta, with mixed success. We also tried hiding the drug in high-value food items such as prunes, apricots, apples, and bananas. Again, the results were mixed. Some animals would initially eat their full dose, but as the study progressed, they were less likely to consume a full dose. The animals never adapted well to the taste and refused all drug-laced food items. To ensure every animal received their appropriate dose, we sometimes had to administer the drug directly into the mouth via gavage. We assume that if the macaques ate most or all of the treated food item, then they would have ingested the full dose of 3 mg/kg.

Control animals were not reluctant to consume their treat, which contained water rather than carpromorelin. Control animals received the same type of treat on any given day in order provide all animals with a similar number of calories. One factor we could not account for was the need to gavage animals if they refused to eat food containing the drug. In those instances, control animals would theoretically ingest more calories on those particular days, at least from the standpoint of treats. However, this is likely a negligible factor because control animals lost weight. We found a statistically significant correlation between capromorelin administration and weight gain. Perhaps other capromorelin compounds would be more palatable to rhesus macaques. This avenue could be explored in future studies.

Additional limitations of this study include small sample size and short duration of treatment. Future studies could use larger cohorts and a wider age range. The mean difference in weight at 14 d between treated and control groups was small (approximately 0.4 kg). A longer dosing period would be needed to determine whether body weight would continue to increase as compared with control macaques. The data from this study can be used for power analyses to determine the samples sizes needed to detect differences in weight and IGF-1 levels between the 2 treatment groups. Due to its short duration, our study is not relevant to the health effects long-term capromorelin has on otherwise healthy macaques.

In summary, we found that capromorelin increased body weight and IGF-1 levels in healthy rhesus macaques over 7 d of treatment period without evidence of any overt negative effects on health. The results from this study suggest that capromorelin may be useful for promoting weight gain in rhesus macaques. Its efficacy and safety should be further studied in larger cohorts over a longer time period. Additional studies in other nonhuman primate species are also warranted.