Anamorelin in the Management of Cancer Cachexia: Clinical Efficacy, Challenges, and Future Directions

Ghrelin and Anamorelin

Ghrelin is a secreted gastric peptide hormone discovered by Kojima and Kamikawa in 1999 (21). It is a natural ligand for growth hormone (GH) secretagogue receptors and has anabolic effects via a transient increase in growth hormone (22). Ghrelin is produced in the stomach (gastric mucosa) and is closely associated with appetite regulation. Gene levels increase during anorexia, fasting, and starvation (23). A review by Prommer summarized the physiological effects of ghrelin on anorexia/cachexia (24). Because ghrelin is a parenteral formulation with a short half-life of approximately 30 minutes, patient burden is high and it is difficult to administer in daily clinical practice.

Anamorelin is an oral drug with ghrelin-like effects. In preclinical studies (in vitro and in vivo studies), anamorelin was shown to be a potent and highly specific ghrelin receptor (growth hormone secretagogue receptor type 1a) agonist with significant effects in stimulating appetite, increasing food intake and weight, and stimulating GH secretion (25) (Figure 1). GH stimulates the secretion of insulin-like growth factor-1 (IGF-1) from the liver, and IGF-1 increases muscle mass. Anamorelin has high affinity (0.70 nM) for ghrelin receptors, which is slightly lower than that of natural ghrelin, and no antagonistic properties (25). An in vitro report which included anamorelin revealed that central access of ghrelin ligands, particularly to the reward areas of the brain, is important for eliciting more potent appetite stimulant effects (26). A study evaluating the effects of ghrelin’s C-terminal portion on pharmacokinetics and biological activity in rats showed that ghrelin promotes GH release primarily via vagal pathways, whereas anamorelin increases GH release via a direct effect on the brain (27). Concern has been expressed that increasing GH and IGF-1 in cancer patients may favor tumor growth, but a study in a mouse NSCLC xenograft model demonstrated that neither anamorelin nor ghrelin promoted tumor growth, despite a trend toward increased mouse GH and mouse IGF-1 (28). In other words, the secretagogue effect of anamorelin on physiologically present GH and IGF-1 is unlikely to promote tumor growth. A prospective study to determine the effect of anamorelin on the endocrine system in patients with gastrointestinal cancer and cachexia reported that levels of thyroid-stimulating hormone (TSH) and free testosterone were altered at 3 weeks after anamorelin administration, but tended to return to equilibrium on administration for more than 12 weeks (29). Other potential effects of anamorelin on the endocrine system are largely unexplored.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Action mechanism of anamorelin (74). GH: Growth hormone; GHRH: growth hormone-releasing hormone; GHS-R_1a: growth hormone secretagogue receptor type 1a; IGF-1: insulin-like growth factor-1.

In healthy volunteers given repeated oral administration of anamorelin (100 mg), the median time to reach maximum plasma concentration (Tmax) was 0.88 (minimum 0.50, maximum 1.5) hours and the maximum plasma concentration (Cmax) was 629±203 ng/ml (30). The Cmax and area under the plasma concentration-time curve from time zero to infinity (AUCinf) of anamorelin (50 mg) administered at 2 hours after the end of a meal were 0.31- and 0.49-fold lower, respectively, than those at fasting. Thus, an effect of diet on Cmax and AUCinf was observed. Anamorelin is mainly metabolized in the liver via CYP3A4, and CYP3A4 inhibitors such as ketoconazole accordingly increase anamorelin Cmax and AUCinf approximately 3-fold (30). Renal dysfunction is expected to have little effect on the pharmacokinetics of anamorelin.

A recent study on the efficacy of anamorelin used prophylactically against the adverse effects of gemcitabine and cisplatin combination chemotherapy in mice (31) reported that it attenuated chemotherapy-induced skeletal muscle atrophy and gastric damage via the downregulation of FOXO1/atrogin-1 signaling. Another report described an inhibitory effect of anamorelin on cisplatin-induced emesis in ferrets, suggesting that brain penetration is important for its anti-emetic effect (32). Nevertheless, the association between anamorelin efficacy and chemotherapy-induced adverse events in humans remains unclear.

Clinical Trials of Anamorelin

Clinical trials evaluating clinical doses of anamorelin (100 mg) are shown in Table III. Most studies administered anamorelin for 12 weeks. In contrast, diagnostic criteria for cancer cachexia (inclusion criteria) and endpoints for evaluating efficacy were not always consistent across trials.

Clinical Trials in Japan. ONO-7643-03 (33) and -04 (34) were phase II studies in Japanese patients with non-small cell lung cancer who have cancer cachexia. In the ONO-7643-03 study, patients were randomized to placebo, or a 50-mg anamorelin or 100-mg anamorelin group. Primary endpoints were mean changes from baseline in LBM, measured by dual energy X-ray absorptiometry (DEXA) and grip strength of the non-dominant hand. Change in LBM over 12 weeks was 0.55±0.29 kg and 1.15±0.31 kg in the placebo and 100-mg anamorelin groups, respectively (p=0.0516), while change in hand grip strength (HGS) was 0.45±0.62 and 1.07±0.67 kg (p=0.3597). In the present study, 100 mg anamorelin resulted in a statistically significant improvement in LBM compared to placebo but no change in HGS of the non-dominant hand. This is because the ONO-7643-03 study reported that 50 mg anamorelin did not significantly improve most of the endpoint variables compared to placebo. The ONO-7643-04 study was designed to evaluate the efficacy of 100 mg anamorelin versus placebo in Japanese patients with non-small-cell lung cancer (NSCLC). The 12-week LBM change was 1.38±0.18 kg in the 100-mg anamorelin group, which was statistically significant compared to −0.17±0.17 kg in the placebo group (p<0.0001). Although the secondary endpoints HGS and 6-minute walk test (6MWT) were measured, no changes between the two groups were observed, indicating no effect on motor activity.

Subsequently, the ONO-7643-05 (35) single-arm study examined the efficacy and safety of 100 mg anamorelin in Japanese patients with gastrointestinal (colorectal, gastric, pancreatic) cancer. Primary endpoint was the proportion of patients classified as LBM responders, namely as patients who maintain or gain LBM (≥0 kg) from baseline to all evaluation time points. Least squares mean±SE change in LBM from baseline was 1.89±0.36 kg, and LBM response rate was 63.3% [95% confidence interval (CI)=48.3-76.6%]. Furthermore, these three clinical trials assessed quality of life using the Quality-of-Life Questionnaire for Cancer Patients Treated With Anticancer Drugs (QOL-ACD). In these studies, 100 mg of anamorelin significantly improved the scores for items 8 “Did you have a good appetite?” and 9 “Did you enjoy your meals?” compared with placebo.

However, the ONO-7643-06 (36) study evaluated the efficacy of anamorelin in patients with cancer and a BMI of less than 20 kg/m² and a weight loss of at least 2% in the last six months, with appetite as primary endpoint. Cancer-related anorexia was quantified using the Functional Assessment of Anorexia/Cachexia Therapy 5-item Anorexia Symptom Scale (FAACT-5ASS) and the FAACT anorexia/cachexia-specific subscale. Eligible patients were administered 100 mg of anamorelin for 24 weeks. Primary endpoint was the composite clinical response (CCR) rate at nine weeks, defined as the percentage of surviving patients who had gained at least 5% body weight from baseline plus an increase of at least 2 FAACT-5IASS score points. CCR rate was 25.9% (95%CI=18.3-35.3%), and 29.0% (95%CI=20.1-39.7%) for patients with NSCLC and 14.3% (95%CI=5.0-34.6%) for those with gastrointestinal cancer. Although a post hoc analysis of the ROMANA 1 and ROMANA 2 trials (described below) did not include CCR rate as a prespecified endpoint, CCR rate at nine weeks was 21.2% and 5.9% in the anamorelin and placebo groups, respectively, among patients with cachexia with BMI <20 kg/m².

Based on efficacy data from these clinical trials, anamorelin received marketing approval in Japan for cancer cachexia in patients with NSCLC and gastrointestinal cancer.

International clinical trials. The ROMANA 1 and ROMANA 2 (37) studies were randomized, double-blind, placebo-controlled phase III trials conducted in 15 and seven countries, respectively, including countries in North America, West Europe and East Europe, as well as Russia and Australia. The co-primary endpoints were change in LBM and HGS over the 12-week study period. Symptoms of anorexia–cachexia and fatigue, measured with the anorexia–cachexia and fatigue scales from the Functional Assessment of Cancer Therapy (FACT) Measurement System, were assessed as part of the secondary endpoints. Mean changes in LBM were +0.99 kg (95%CI=0.61-1.36 kg) in the anamorelin 100 mg group and −0.47 kg (95%CI=−1.00-0.21 kg) in the placebo group (p<0.0001) in the ROMANA 1 study, and +0.65 kg (95%CI=0.38-0.91 kg) and −0.98kg (95%CI=−1.49 − −0.41 kg) (p<0.0001) in the ROMANA 2 study, respectively. Even though no statistically significant difference in mean change in HGS and fatigue scale was observed, mean anorexia–cachexia scale score was significantly improved in the 100-mg anamorelin groups in both studies. Subsequently, the ROMANA 3 (38) study was conducted as a safety extension study of the ROMANA 1 and 2 studies to assess the safety and efficacy of anamorelin in enrolled patients completing the 12-week anamorelin regimen in the original trials. The most frequent adverse event was hyperglycemia, which occurred in 1.2% of the 100-mg anamorelin group versus 0% in the placebo group. Anamorelin was well tolerated, showing an adverse event profile similar to that in the ROMANA 1 and 2 studies. The 100-mg anamorelin group showed significantly improved body weight over 24 weeks versus the placebo group (3.1±0.6 kg; 95%CI=1.8-4.3 kg vs. 0.9±0.7 kg; 95%CI=−0.5-2.3 kg). In contrast, no improvement in HGS was observed in either group. From these results, it was determined that anamorelin has limited effect on improving LBM and no effect on HGS or QOL, and was not approved in Europe.

Clinical trial endpoints. Four meta-analyses of randomized trials have evaluated the efficacy of anamorelin (39-42). Some of the clinical trials analyzed in these meta-analyses included a 50-mg anamorelin arm. In three of the meta-analyses, LBM, body weight and QOL were shown to have improved significantly in the anamorelin group compared with the placebo group, whereas there was no difference in HGS or overall survival. As noted above, while clinical trials of anamorelin have reported approximately consistent trends in efficacy, decisions on whether or not to approve this agent differ internationally. Expanding clinical use requires the demonstration of benefits in both behavioral change and in the survival of patients with cancer cachexia. A systematic review of treatments for cancer cachexia, including anamorelin, mentioned that there is an urgent need to rethink what is “clinically appropriate” in the study and treatment of cancer cachexia and to meet the demands of patients, researchers, and regulatory authorities; and that simultaneous improvement in skeletal muscle mass, physical function, QOL, and overall survival may be an “ideal” endpoint (43). Similarly, a working consensus report from the AWGC concluded that the crucial outcomes for Asian patients with cachexia were mortality, QOL, and functional status (2). In other words, the true goal of anamorelin should be defined as improvements in these endpoints.

Previous clinical trials have generally used LBM as primary endpoint to evaluate the efficacy of anamorelin, while HGS and 6-minute walk have been measured as indicators of behavioral change. However, previous clinical trials of the efficacy of drugs for cancer cachexia have not always assessed motor function. Several of the clinical trials which did assess motor function used HGS, 6-minute walk, stair climb time, performance status (PS), and functional areas of QOL assessment (44-48). In a systematic review that assessed the frequency and diversity of physical function endpoints in cancer cachexia trials, HGS was shown to be the most commonly used physical function endpoint (49). However, heterogeneity in clinical trials makes it difficult to identify an optimal endpoint. However, endogenous substances assayed from blood and related to inflammation, such as albumin and C-reactive protein, are frequently used as biomarker endpoints in cancer cachexia trials (50). Many of these biomarkers are not directly related to the mechanism of anamorelin, however, and their use in routine practice has not been established.

Selection of optimal target patients. Post-hoc analyses of the ONO-7643-04 study were performed to investigate the efficacy and safety of anamorelin in subgroups categorized by sex, age, body mass index, body weight loss within the last 6 months, ECOG PS, concomitant anticancer therapy, number of prior chemotherapy regimens, C-reactive protein, albumin, and hemoglobin (51). Change in LBM from baseline to week 12 was significantly greater with anamorelin than placebo in almost all patient subgroups, except in the PS 2 subgroup. In addition, the least-squares mean of LBM for anamorelin was greater in patients with fewer prior chemotherapy regimens; accordingly, this post-hoc analysis suggested that anamorelin may be more effective when used precociously. A notable result in subgroup analyses of the ROMANA studies was that the male subgroup in ROMANA 1 was the only subgroup with a statistically significant treatment difference in change in HGS from baseline over 12 weeks between anamorelin and placebo (p=0.024) (37). For other subgroup data, the studies were generally similar.

One preceding review (20) points out that one of the problems related to clinical trials of anamorelin is the dual incorporation of patients in the early “pre-cachexia” phase and those in the later “refractory cachexia” phase. Refractory cachexia is unresponsive to any medication. In an interim analysis of post-marketing surveillance in Japan, treatment with anamorelin was discontinued in 22.1% of patients due to poor response (52). The definition of cancer cachexia in Japanese clinical trials was involuntary weight loss >5% within the last six months, anorexia, two or more applicable symptoms (fatigue, malaise, reduced overall muscular strength, and arm muscle circumference <10^th percentile), and one or more of the following serum biomarkers: albumin level <3.2 g/dl, C-reactive protein level >5.0 mg/l, and hemoglobin level <12 g/dl (34, 35). The ROMANA studies defined cancer cachexia as involuntary weight loss of ≥5% within the previous 6 months, or body-mass index <20 kg/m² (37). These diagnostic criteria are based on international guidelines (4, 53), however, and there are currently no clinical data for anamorelin based on the AWGC diagnostic criterion for Asian patients. Patient selection based on this new diagnostic criterion may reveal a benefit of anamorelin.

Table IV summarizes reports on clinical outcomes of anamorelin in real world practice. There are scattered retrospective studies in patients with pancreatic cancer. In an interim analysis of post-marketing surveillance in Japan, anamorelin was the second-most frequently administered drug to patients with pancreatic cancer, following those with NSCLC (52), but a single-center study reported that anamorelin usage was most predominant for those with pancreatic cancer (54). In a retrospective study (55) in 20 patients with advanced gastrointestinal cancer, clinical characteristics were compared between anamorelin responders and non-responders. Identified nutritional markers such as total protein, albumin, transferrin, and prognostic nutritional index at baseline were significantly higher in responders, whereas systemic inflammatory markers such as neutrophil/lymphocyte and C-reactive protein/albumin ratios at baseline were significantly higher in non-responders. Another retrospective study in 24 patients with pancreatic cancer compared the efficacy of anamorelin in moderate-weight-loss (5-10%) and severe-weight-loss groups (>10%) (56). Although appetite was consistently improved in both groups, weight gain was significantly greater in the moderate-weight-loss group than in the severe-weight-loss group. It should be noted, however, that anamorelin was discontinued in 22 patients (7/8 in the moderate-weight-loss group and 15/16 in the severe-weight-loss group) during the study period, and only three patients in the moderate-weight-loss group completed 24 weeks of anamorelin. Twelve of 24 patients discontinued anamorelin because of clinical deterioration. Another study which evaluated the efficacy of anamorelin included patients with PS 2 (57), and found that the proportion of responders was significantly lower in patients with poor PS than in those with good PS (0% vs. 37%, p=0.042). The C-reactive protein-albumin ratio (CAR), an indicator of systemic inflammation, is reported to be a risk factor for nonresponse to anamorelin (odds ratio= 5.6, 95%CI=1.2-27.1, p=0.032) (58), suggesting that there is a close association between the exertion of anamorelin’s action and inflammatory conditions. The introduction of anamorelin at an earlier phase of cancer cachexia may therefore be the key to ensuring efficacy, and should be validated in future studies.

Quality of life assessment tools. Clinical trials of anamorelin employ QOL assessment tools to evaluate anorexia and other symptoms associated with cancer cachexia. The clinical trials in Japan (33-36) focused on items 8, 9, and 11 of the QOL Questionnaire for Cancer Patients Treated with Anticancer Drugs (QOL-ACD). These items assess patient complaint-based symptoms mainly related to appetite and weight using the following questions: item 8 “Did you have a good appetite?”, item 9 “Did you enjoy your meals?”, and item 11 “Did you lose any weight?”. The QOL-ACD consists of four domains (functional, physical, spiritual, and psychosocial) and a global face scale, and a previous report suggested that it may be useful for clinical studies in Japanese patients with advanced NSCLC (61). In the ROMANA studies (37, 38), symptom burden was quantified with the 12-item anorexia-cachexia scale of the FAACT and the 13-item fatigue scale of the Functional Assessment of Chronic Illness Therapy-Fatigue (FACIT-F). FAACT is composed of the FACT Measurement System scale and the anorexia–cachexia scale, while FACIT-F is composed of the FACT scale and the fatigue scale. Both tools have been validated for their application to patients with cancer (6). An additional analysis of the ROMANA 2 study was performed to identify the more responsive of the two QOL measures (63). The results indicated that the fatigue/activity and appetite/eating scales derived from the FACIT-F and FAACT were reliable, valid, and responsive. The use of these quality-of-life measures in clinical practice may assist in evaluating the efficacy of anamorelin and in making decisions about its continuation in patients with cancer cachexia.

Nutrition and exercise therapy. Although treatment strategies for cancer cachexia require a multidisciplinary approach, none of the clinical trials of anamorelin has specifically mentioned nutritional or exercise therapy interventions (33-38). As we previously reported, we are practicing a management system for anamorelin treatment that involves the collaboration of physicians, nurses, pharmacists, and dietitians (64).

Some studies have reported the use of nutritional assessment indicators to explore predictors of anamorelin efficacy. Fujita et al. found that mGPS 2 was a risk factor, which affected the 12-week continuation rate of anamorelin (60). Furthermore, we previously reported that a low Controlling Nutritional Status (CONUT) score was significantly associated with “12-week sustained effective response” to anamorelin treatment in multivariable logistic analysis (61). Assessment of nutritional status can be an important factor in validating the efficacy of anamorelin.

Anamorelin is a stimulator of GH secretion (25). GH in turn increases muscle mass but not necessarily muscle strength (65). In other words, muscle mass is not directly related to muscle quality (66), and treatment with anamorelin alone may not achieve sufficient improvement in muscle strength. Further, patients with cancer cachexia had a decrease in stair-climbing ability but not in muscle radiodensity (reflecting muscle quality), indicating that muscle mass, but not radiodensity, predicts functional performance in patients with cancer (67). However, the utility of a combination of anamorelin treatment and exercise therapy has not been reported.

In a prospective cross-sectional study of patients with cancer-related fatigue (FACIT-F ≤34), anamorelin and physical activity and nutrition counseling were combined for 43 days to assess changes in cancer-related fatigue (68). An association with improvement in fatigue was detected using multiple measures; however, no significant improvement in physical performance outcomes was found, despite improvements in weight and body composition. A report examining potential causes for the lack of association between muscle mass and physical function in cancer cachexia (69) mentioned that the relationship between muscle mass and muscle function is complex and is influenced by the technique used to assess the nutritional endpoint, the nature of the selected physical function outcome measures, and the sex and severity of the recruited patients with cachexia. A multinational randomized phase II trial of multidisciplinary interventions in patients with NSCLC and pancreatic cancer (Pre-MENAC trial, feasibility study) evaluated interventions using nutrition counseling, aerobic and resistance training, celecoxib, and energy-dense supplements high in eicosapentaenoic acid (70). Compliance with combinations of two interventions ranged from 20% to 48%, while compliance with all three interventions was 12%, suggesting a trade-off between the number of interventions and compliance. Therefore, any exploration of combination therapies, including anamorelin, that are effective in patients with cancer cachexia must consider that treatment outcomes may be affected by the selection of target patients, the burden of treatment on patients, and the assessment tools used by healthcare providers. Further clinical trials of the efficacy of multimodal treatments for cancer cachexia are currently underway (71, 72), and the results are awaited.

Adverse events. Comparatively frequent major adverse events in clinical trials and an interim analysis of post-marketing surveillance of anamorelin included hyperglycemia, increases in glycosylated hemoglobin and γ-glutamyltransferase, nausea, first-degree atrioventricular block, and prolonged QRS complex (a measure of ventricular depolarization on electrocardiograms, ECG) (32-37, 51). In the ROMANA 3 study, tolerability of anamorelin over 24 weeks was demonstrated (38).

As would be expected with increased appetite and food intake due to anamorelin, even in retrospective data, the most common adverse event leading to treatment discontinuation of anamorelin in patients with advanced pancreatic cancer and diabetes was hyperglycemia (56). In one case report, poorly controlled hyperglycemia developed even with increasing doses of insulin, raising concerns about the possibility that anamorelin induces peripheral insulin resistance (73). In a retrospective study of the efficacy and safety of anamorelin in patients with diabetes, there were fewer responders in the diabetic group than in the non-diabetic group (45.2% vs. 81.9%, p<0.01), a higher cumulative incidence of hyperglycemic adverse events (72.2% vs. 6.3%, p<0.01), and more discontinuations due to adverse events (25.8% vs. 4.2%, p<0.01) (59). One report used a Japanese claims database to examine risk factors of adverse metabolic effects on glucose level induced by anamorelin, with a focus on sex (male), age (>75 years), types of carcinoma, history of diabetes mellitus, and concomitant use of steroids as factors. Results showed that pancreatic cancer and history of diabetes were associated with glucose metabolism disorder (74). In patients with cancer cachexia, assessment of glycemic control and insulin resistance may be useful from the time of anamorelin initiation.

In several clinical trials in which antiarrhythmic drugs were not permitted, the occurrence of arrhythmias may have been attributable to inhibition of the stimulus conduction system due to anamorelin’s sodium channel blockade effects (25). Several case reports have described the onset of arrhythmia under anamorelin treatment with severe outcomes (75-77). Common features in these patients include hepatic dysfunction and hypoproteinemia. Given that anamorelin is mainly metabolized in the liver via CYP3A4 and has a high protein binding rate (97.3-98.3%) (30), blood levels of anamorelin may be elevated in patients with these comorbid conditions. However, one case report described fatal arrhythmia complicated by sinus arrest and refractory ventricular tachycardia, despite the absence of severe hepatic dysfunction and CYP3A4 inhibitor (78). The possibility that arrhythmias induced by anamorelin are multifactorial must be kept in mind. Because of the lack of uniform ECG monitoring in clinical trials, arrhythmias and abnormal ECG may not have been appropriately detected (79).

In a prospective observational study of multimodal clinical practice, early discontinuation (less than four weeks) of anamorelin was associated with worse PS and low prognostic nutritional index (80). The reason for discontinuation in 47.4% of patients in the early discontinuation group was adverse effects, including hyperglycemia, diarrhea and cardiac conduction disturbances. Management of adverse effects induced by anamorelin may affect its effect in friable patients. One reason for the refusal of the European Medical Agency (EMA) to provide marketing authorization for anamorelin was that the safety data had not been recorded adequately, and a thorough evaluation of potential risks was accordingly not possible (81). Attention should be paid to the above conditions and concomitant medications (especially CYP3A4 inhibitors), and future efforts should be made to accumulate and analyze safety data.

The Future of Anamorelin

The treatment of cancer cachexia requires a multi-disciplinary approach consisting of nutritional intervention, physical exercise, and pharmacotherapy. Clear evidence for combination therapy, which includes anamorelin is needed in the future. In addition, some reports suggest that multiple drug combinations may be necessary for anamorelin therapy, such as targeting Activin A-induced metabolic dysfunction (82), and basic research into the development of new drugs for use in combination with anamorelin is in fact underway (83). Optimizing the use of anamorelin holds the promise of improving quality of life and survival outcomes for patients with cancer cachexia.