Effects by daily long term provision of ghrelin to unselected weight-losing cancer patients
A randomized double-blind study
Abstract
BACKGROUND:
The short-term provision of ghrelin to patients with cancer indicates that there may be benefits from long-term provision of ghrelin for the palliative treatment of weight-losing cancer patients. This hypothesis was evaluated in a randomized, double-blind, phase 2 study.
METHODS:
Weight-losing cancer patients with solid gastrointestinal tumors were randomized to receive either high-dose ghrelin treatment (13 μg/kg daily; n = 17 patients) or low-dose ghrelin treatment (0.7 μg/kg daily; n = 14 patients) for 8 weeks as a once-daily, subcutaneous injections. Appetite was scored on a visual analog scale; and food intake, resting energy expenditure, and body composition (dual x-ray absorpitometry) were measured before the start of treatment and during follow-up. Serum levels of ghrelin, insulin, insulin-like growth factor 1, growth hormone (GH), triglycerides, free fatty acids, and glucose were measured. Health-related quality of life, anxiety, and depression were assessed by using standardized methods (the 36-item Short Form Health Survey and the Hospital Anxiety and Depression Scale). Physical activity, rest, and sleep were measured by using a multisensor body monitor.
RESULTS:
Treatment groups were comparable at inclusion. Appetite scores were increased significantly by high-dose ghrelin analyzed both on an intent-to-treat basis and according to the protocol. High-dose ghrelin reduced the loss of whole body fat (P < .04) and serum GH (P < .05). There was a trend for high-dose ghrelin to improve energy balance (P < .07; per protocol). Otherwise, no statistically significant differences in outcome variables were observed between the high-dose and low-dose groups. Adverse effects were not observed by high-dose ghrelin, such as serum levels of tumor markers (cancer antigen 125 [CA 125], carcinoembryonic antigen, and CA 19-9).
CONCLUSIONS:
The current results suggested that daily, long-term provision of ghrelin to weight-losing cancer patients with solid tumors supports host metabolism, improves appetite, and attenuates catabolism. Cancer 2010. © 2010 American Cancer Society.
Anorexia and weight loss are severe health-related quality of life problems for patients with progressive malignant disease. Recent studies in our laboratory have indicated that systemic anti-inflammatory treatment, nutritional support, and insulin provision to cachectic cancer patients may all improve nutritional status, health-related quality of life, and survival in unselected patients.1-3 These findings are encouraging and demonstrate that a variety of metabolic interventions can improve well being and physical functioning in palliative settings.4, 5 However, anorexia has been particularly difficult to counteract consistently despite well described concepts, although short-term benefits have been reported from trials with steroid interventions to cancer patients.6, 7 Therefore, a possibility to treat cancer patients with orexigenic drugs should benefit a large group of patients.8, 9 Results from acute or short-term provision of ghrelin or ghrelin agonists to healthy volunteers and patients with cancer support the idea of benefits from long-term provision of ghrelin for palliative support in weight-losing cancer patients.10-13 Therefore, the objective of the current study was to evaluate the long-term effects of giving daily synthetic ghrelin to unselected, weight-losing cancer patients in a randomized, double-blind study (National Clinical Trial no. NCT00681486).
MATERIALS AND METHODS
Patients with systemic and progressive cancer were invited to participate in our study for inclusion between 2006 and 2008 at our out patient clinic (Fig. 1). Inclusion criteria were progressive gastrointestinal cancer with systemic spread without any further, ongoing specific tumor therapy; weight loss >5% and subjective appreciation of anorexia as a main cause behind health-related problems. Exclusion criteria were age <40 years, manifest diabetes (type I or II), steroid treatment, or any other treatment known or assumed to affect tumor host metabolism. Patients who were receiving nutritional support of any kind were not invited to participate.
This chart illustrates the start and continuation of high-dose and low-dose ghrelin treatment, as described in the text (see Material and Methods), and inclusion and follow-up schema for patients who received high-dose and low-dose ghrelin treatment as daily subcutaneous injections before the main daily main meal. QoL indicates quality of life; SF-36, 36-item Short Form Health Survey; HAD, Hospital Anxiety and Depression Scale; DEXA, dual x-ray absorpitometry.
Treatment protocol, prerequisites and expectations were explained for candidate patients. Inclusion measurements, which agreed with outcome variables, were secured when patients agreed to participate. All patients were randomized for either high-dose ghrelin treatment (∼10 μg/kg daily) or low-dose ghrelin treatment (0.5 μg/kg daily). A trained nurse was responsible for information and medical care interactions with patients' medical home care teams for treatment instructions at start up. All patients were followed by a home care team. Some patients subsequently asked for education, because they wanted to take responsibility for their daily subcutaneous injections according to similar, routine procedures.3 Patients and home care providers were instructed to give ghrelin injections 30 to 45 minutes before the main daily meal, which usually occurred between 11 AM and 12 noon All injections were double blind. Ghrelin injections were scheduled to continue daily for 8 weeks with outcome variables registered before and after this period. Some measurements also were obtained after 4 weeks of treatment in combination with a general clinical evaluation by the physician in charge (Fig. 1). Twelve patients who were receiving high-dose ghrelin and 10 patients who were receiving low-dose ghrelin finished the intended 8 weeks of treatment. Interruption of treatment was either for personal (private) reasons (3 patients who were receiving high-dose ghrelin and 2 patients who were receiving low-dose ghrelin) or because of progressive disease (2 patients who were receiving high-dose ghrelin and 2 patients who were receiving low-dose ghrelin).
Measurements at Inclusion and Follow-Up
Measurements for use in stratified randomization and follow-up were tumor type, tumor stage, previous tumor treatment (surgery, chemotherapy, or radiotherapy), age, sex, height, nutritional status (body weight, weight loss, serum albumin concentration), liver function tests (aspartate aminotransferase, alanine aminotransferase, and alkaline phosphatase levels), serum creatinine, blood hemoglobin concentration, erythrocyte sedimentation rate, and C-reactive protein, and any previous use of analgesics, beta blockers, or nonsteroidal anti-inflammatory drugs. Measurements were repeated at 4 weeks and 8 weeks after inclusion (Fig. 1). Physiologic variables at rest included heart rate, systolic and diastolic blood pressure, body temperature, respiratory rate, and resting energy expenditure. All drug medications were assessed before and during follow-up.
Nutritional Assessments
Food intake was assessed by a dietician according to 4-day records, as described previously.14, 15 Urine samples were used to verify registered protein intake. Resting energy balance was derived as energy intake minus resting energy expenditure to reflect the amount of energy available for physical activity and repletion of energy stores. Body composition was measured by dual energy x-ray absorptiometry as described elsewhere.2, 3, 14, 16 Indirect calorimetry was used to determine resting energy expenditure (Deltatrac; Datex, Helsinki, Finland) in the morning after an overnight fast as described elsewhere.14, 17
Blood Tests
Blood chemistry analyses included hemoglobin, glucose, insulin, C-peptide, insulin-like growth factor 1 (IGF-1), growth hormone (GH), triglycerides (TG), free fatty acids (FFA), white blood cell and thrombocyte counts, erythrocyte sedimentation rate, C-reactive protein, albumin, electrolyte and creatinine levels, liver function tests, and tumor markers (carcinoembryonic antigen [CEA], cancer antigen 19-9 [CA 19-9], and CA 125) measured in the Department of Clinical Chemistry at our hospital. Total ghrelin was measured by using a radioimmunoassay according to the manufacturer's instructions (GHRT89HK; Millipore/Linco, Billerica, Mass), and active ghrelin was measured by using an enzyme-linked immunosorbent assay (EZGRA8K; Millipore/Linco).
Health-Related Quality of Life
Health-related quality of life was assessed by using 1 global instrument (the 36-item Short Form Health Survey [SF-36]) and 1 anxiety and depression scale (the Hospital Anxiety and Depression Scale [HADS]).18, 19
Daily Physical Activity
Spontaneous daily activity was measured for 3 days by the Sense Wear System (Body Media, Pittsburgh, Pa; available at: www.bodymedia.com accessed on August 30, 2008). The Bodymedia SenseWear Pro 2 armband is a multisensor body monitor worn on the triceps of the arm. It can be worn for days or weeks and continuously records physical activity signals. Periods and levels of physical activity are quantified while the monitor measures the number of steps, movements, and sleep and wake periods. Physical activities are translated into energy expenditure by an exclusive algorithm provided by the manufacturer.
Ghrelin
Ghrelin was provided free of charge by Gastrotech Pharma A/S (Copenhagen, Denmark). The compound was synthesized by Polypeptide Laboratories (Wolfenbuttel, Germany). The peptide content of the substance was 84.3% with acetate as the counter ion. The preparations had between 98.41% and 99.63% peptide purity (catalog no. P-0861, lot no. H-a-1096). The substance was prepared using guanylate, was dissolved in physiologic saline (0.9%) on the day of injection, and was brought to the patient by a nurse who administered the drug. Patients received ampoules that contained either 800 μg or 40 μg corresponding to high-dose and low-dose ghrelin, respectively, according to randomization. The amounts provided corresponded to 13 ± 1 μg/kg daily or 0.7 ± 0.04 μg/kg daily to account for individual body weight based on 56 injections for each group over 8 weeks.
Statistics
Patients were randomized by a computerized algorithm to high-dose ghrelin (n = 17) or low-dose ghrelin (n = 14) by stratification as described by Pocock and Simon.20 The following stratification variables were included to reduce the risk of inappropriate skewness in important background variables: sex, age, body mass index (BMI), weight loss, serum albumin, erythrocyte sedimentation rate, C-reactive protein, liver function tests, drug intake, tumor type, and disease stage. Power estimates before the study on appetite scoring (α = .05, β = .80) made it likely to detect differences between 31 randomized cancer patients according to our unpublished experience. We hypothesized that elevated appreciation of food should be translated into statistically significant nutrition outcome variables. The results are presented as mean ± standard error of the mean. All assessments of high-dose and low-dose treatment were analyzed by using an analysis of variance (ANOVA) for repeated measures. Two group analyses were conducted using a 1-way ANOVA. P values <.05 were considered statistically significant, and P values <.10 were considered a trend toward significance in 2-tailed tests. All results were evaluated primarily on an intent-to-treat basis and then according to the protocol (high-dose group, n = 12 patients; low-dose group, n = 10 patients in all tables and figures). This study was approved by the Regional Ethical Review Board, Gothenburg (S543-03).
RESULTS
Table 1 provides information on clinical characteristics and laboratory test values at inclusion for all randomized patients. Statistically significant differences were not observed between patients in the high-dose and low-dose groups at inclusion when they were analyzed on an intent-to-treat basis or according to the protocol other than anamnestic weight loss (P < .04). Food intake and energy expenditure did not differ significantly between patients in the high-dose or low-dose group over time. However, overall, subjectively scored appetite was significantly higher in patients who received high-dose ghrelin compared with patients who received low-dose ghrelin (Table 2, Fig. 2). Serum albumin, insulin, and IGF-1 levels as well as serum TG, FFA, and glucose did not differ between the 2 groups at inclusion or at the end of 8 weeks of treatment (Table 2). Serum ghrelin did not differ significantly among the treatment groups at inclusion or over time, whereas basal serum GH levels decreased across 8 weeks of treatment among patients who were receiving high-dose ghrelin. Two patients who were receiving low-dose ghrelin spontaneously had very high levels of both active and total serum ghrelin. The endogenously increased levels in these patients did not seem to translate into altered clinical characteristics, such as appetite score or food intake (results not shown).
| Total No. of Patients (No. Who Completed Treatment) | ||
|---|---|---|
| Cancer Site | High-Dose Ghrelin, n=17 | Low-Dose Ghrelin, n=14 |
| Colorectal | 1 (1) | 0 (0) |
| Esophageal/gastric | 9 (7) | 7 (4) |
| Pancreatic | 5 (4) | 2 (2) |
| Liver/bile duct | 2 (0) | 2 (2) |
| Intestinal | 0 (0) | 3 (2) |
| Mean ± SEM | ||
|---|---|---|
| Variable | High-Dose Ghrelin, n=17 | Low-Dose Ghrelin, n=14 |
| Clinical characteristics | ||
| Age, y | 73 ± 2 | 76 ± 2 |
| Men:women | 11:6 | 7:7 |
| Length, cm | 172 ± 2 | 169 ± 4 |
| Weight, kg | 61.5 ± 3.1 | 61.5 ± 3.4 |
| Body mass index, kg/m2 | 21.0 ± 0.8 | 22.0 ± 1.2 |
| Weight loss, % | 15 ± 1a | 20 ± 2 |
| Laboratory test values | ||
| S-creat, μmol/L | 77 ± 6 | 79 ± 9 |
| S-Bi, μmol/L | 19 ± 5 | 12 ± 2 |
| S-alp, μkat/L | 7 ± 3 | 2 ± 1 |
| AST, μkat/L | 1.2 ± 0.4 | 0.6 ± 0.2 |
| ALT, μkat/L | 0.8 ± 0.2 | 0.4 ± 0.1 |
| Hemoglobin, g/L | 126 ± 3 | 122 ± 4 |
| ESR, mm/h | 28 ± 5 | 33 ± 5 |
| CRP mg/L | 26 ± 1 | 22 ± 5 |
| S-alb, g/L | 32 ± 1 | 31 ± 2 |
| P-glucose, mmol/L | 6.1 ± 0.2 | 6.8 ± 0.5 |
| S-IGF-1, μg/L | 79 ± 11 | 82 ± 11 |
| CA 125, U/L | 211 ± 75 | 238 ± 177 |
| S-insulin, mUnits/L | 18 ± 6 | 12 ± 3 |
| S-TG, mmol/L | 1.26 ± 0.13 | 1.03 ± 0.08 |
| S-FFA, mmol/L | 0.51 ± 0.07 | 0.52 ± 0.06 |
| REE, kcal | 1451 ± 59 | 1343 ± 65 |
| REE/kg, kcal/kg | 29 ± 4 | 24 ± 2 |
| Syst BP, mm Hg | 133 ± 5 | 132 ± 7 |
| Diast BP, mm Hg | 77 ± 3 | 73 ± 3 |
| Heart rate, beats/min | 74 ± 3 | 72 ± 3 |
- SEM indicates standard error of the mean; S-creat, serum creatinine; S-Bi, serum bilirubin; S-alp, serum alkaline phosphatase; AST, aspartate aminotransferase; ALT, alanine aminotransferase; ESR, erythrocyte sedimentation rate; CRP, C-reactive protein; S-alb, serum albumin; P-glucose, plasma glucose; S-IGF-1, serum insulin-like growth factor 1; CA 125, cancer antigen 125; S-TG, serum triglycerides; S-FFA, serum free fatty acids; REE, resting energy expenditure; Syst, systolic; BP, blood pressure; Diast, diastolic.
- a Significant differences other than anamnestic weight loss (P < .04) were not observed between patients on high-dose and low-dose ghrelin treatment for this measurement.
| Mean±SEM | |||
|---|---|---|---|
| Variable | At Inclusion | At 8 Weeks | Pa |
| S-ghrelin total, ng/L | |||
| High-dose group | 563 ± 90 | 1229 ± 501 | NS |
| Low-dose group | 3418 ± 2570 | 3817 ± 2997 | |
| S-ghrelin active, ng/L | |||
| High-dose group | 96 ± 30 | 178 ± 72 | NS |
| Low-dose group | 604 ± 336 | 543 ± 338 | |
| S-insulin, mUnits/L | |||
| High-dose group | 22 ± 10 | 29 ± 12b | NS |
| Low-dose group | 16 ± 3 | 11 ± 3 | |
| S-IGF-1, μg/L | |||
| High-dose group | 100 ± 16 | 72 ± 12 | NS |
| Low-dose group | 84 ± 14 | 76 ± 19 | |
| S-GH, μg/L | |||
| High-dose group | 3.1 ± 1.1 | 1.3 ± 0.4 | <.05 |
| Low-dose group | 3.9 ± 1.2 | 6.0 ± 3.4 | |
| S-albumin, g/L | |||
| High-dose group | 32.1 ± 1.7 | 36.1 ± 1.9 | NS |
| Low-dose group | 31.8 ± 1.9 | 28.6 ± 3.2 | |
| S-TG, mmol/L | |||
| High-dose group | 1.26 ± 0.17 | 1.27 ± 0.19 | NS |
| Low-dose group | 0.87 ± 0.06 | 1.07 ± 0.19 | |
| S-FFA, mmol/L | |||
| High-dose group | 0.52 ± 0.10 | 0.61 ± 0.11 | NS |
| Low-dose group | 0.49 ± 0.09 | 0.46 ± 0.07 | |
| S-glucose, mmol/L | |||
| High-dose group | 6.2 ± 0.3 | 6.0 ± 0.4 | NS |
| Low-dose group | 7.3 ± 0.7 | 6.6 ± 0.7 | |
| Energy expenditure, kcal/kg/d | |||
| High-dose group | 24.5 ± 0.9 | 23.5 ± 1.2 | NS |
| Low-dose group | 21.9 ± 0.7 | 22.5 ± 1.3 | |
| Food intake, kcal/kg/d | |||
| High-dose group | 32.5 ± 9.4 | 28.2 ± 3.8 | NS |
| Low-dose group | 24.1 ± 3.0 | 25.5 ± 4.5 | |
| Appetite score, 1-10 | |||
| High-dose group | 5.6 ± 1.0 | 6.8 ± 0.7 | <.02 |
| Low-dose group | 3.9 ± 0.6 | 4.0 ± 1.1 | |
| Whole body fat, kg | |||
| High-dose group | 15.1 ± 1.9 | 13.5 ± 1.9 | −1.3 ± 0.7c |
| Low-dose group | 16.3 ± 3.0 | 12.6 ± 2.4 | −3.7 ± 0.8c |
| Whole body fat free mass, kg | |||
| High-dose group | 44.8 ± 2.9 | 47.8 ± 2.9 | 2.24 ± 0.71d |
| Low-dose group | 44.3 ± 2.2 | 45.1 ± 2.8 | 0.86 ± 1.18d |
- SEM indicates standard error of the mean; S, serum; NS, not significant; IGF-1, insulin-like growth factor 1; GH, growth hormone; TG, triglycerides; FFA, free fatty acids.
- a Analysis of variance for repeated measures.
- b The glucose-to-insulin ratio did not differ among patients in the high-dose and low-dose groups.
- c Values show the difference over time (P < .04).
- d Values show the difference over time (P < .3).
(Top) The appetite visual analog score is illustrated in patients who were randomized to receive high-dose (n = 17) or low-dose (n = 14) ghrelin on an intent-to-treat basis at all available registrations (high-dose ghrelin, n = 42; low-dose ghrelin, n = 30; P < .002). (Bottom) Resting energy balance is illustrated in patients who were randomized to receive high-dose (n = 12) and low-dose (n = 10) ghrelin treatment according to protocol for 8 weeks (P < .07). Patients on high-dose ghrelin were in positive energy balance (P < .05). Calculations were based on all available registrations at inclusion and during follow-up (high-dose ghrelin, n = 25; low-dose ghrelin, n = 20 patients).
Whole-body fat loss was less pronounced in cancer patients who were receiving daily high-dose ghrelin compared with low-dose treatment (P < .04) (Table 2). Whole-body fat-free mass (lean tissue) had a similar pattern but did not reach statistical significance. Improved fat retention in patients who were receiving high-dose ghrelin was reflected by a trend toward improved energy balance during high-dose ghrelin treatment (P < .07) (Fig. 2), because a significantly positive energy balance was observed among patients in the high-dose group (P < .05) (Fig. 2).
Health-related quality of life measured by global scores (SF-36) as well as anxiety and depression scales (HADS) did not differ over time between patients in the high-dose and low-dose groups (Table 3). However, patients who were receiving high-dose ghrelin had significantly less anxiety and higher scores for mental health at inclusion compared with patients who were receiving low-dose ghrelin (Table 3). Physical activity (kcal daily), rest per day (minutes), and sleep per day (minutes) did not differ at inclusion or at 8 weeks of follow-up (Table 4). No objective adverse effects were registered or associated with ghrelin injections. Tumor markers did not increase significantly in patients in the high-dose group (CA 125, from 124 ± 46 U/L to 135 ± 43 U/L [difference not significant]; CEA, from 6 ± 2 μg/L to 9 ± 5 μg/L [difference not significant]; CA 19-9, from 90 ± 28 U/L to 269 ± 145 U/L [difference not significant]). Patients in the high-dose group had lower CA 125 and CA 19-9 levels at inclusion and at follow-up compared with patients in the low-dose group, but the difference between groups was not statistically different over time (results not shown).
| Mean±SEM | ||
|---|---|---|
| Measure | At Inclusion | At 8 Weeks |
| HADS anxiety | ||
| High-dose group | 4.6 ± 1.07a | 5.4 ± 1.6 |
| Low-dose group | 8.6 ± 1.6 | 8.8 ± 1.3 |
| HADS depression | ||
| High-dose group | 5.7 ± 0.9 | 6.8 ± 1.2 |
| Low-dose group | 7.6 ± 1.4 | 9.3 ± 1.9 |
| SF-36 PCS | ||
| High-dose group | 30 ± 3 | 27 ± 3 |
| Low-dose group | 35 ± 3 | 30 ± 2 |
| SF-36 MCS | ||
| High-dose group | 40 ± 3a | 41 ± 3 |
| Low-dose group | 30 ± 2 | 34 ± 4 |
- SEM indicates standard error of the mean; HADS, Hospital Anxiety and Depression Scale; SF-36, 36-item Short Form Health Survey; PCS, physical component scale; MCS, mental component scale.
- a P < .05 versus the low-dose group.
| Mean±SEM | |||
|---|---|---|---|
| Measure | At Inclusion | At 8 Weeks | P |
| Physical activity, kcal/d | |||
| High-dose group | 2025 ± 114 | 2010 ± 137 | NS |
| Low-dose group | 1834 ± 100 | 2208 ± 116 | |
| Physical activity, steps/d | |||
| High-dose group | 1259 ± 75 | 2641 ± 810 | NS |
| Low-dose group | 4929 ± 2286 | 3700 ± 1107 | |
| Rest per d, mina | |||
| High-dose group | 581 ± 38 | 522 ± 53 | NS |
| Low-dose group | 601 ± 24 | 598 ± 39 | |
| Sleep per d, min | |||
| High-dose group | 475 ± 30 | 411 ± 39 | NS |
| Low-dose group | 509 ± 22 | 467 ± 54 | |
- SEM indicates standard error of the mean; NS, not significant.
- a Includes sleep; the mean±SEM of all measurement occasions at inclusion and at 8-week follow up.
DISCUSSION
Anorexia and malnutrition are severe health-related quality-of-life–related sequelae in patients with progressive cancer disease.6 Several previous attempts have been made to attenuate anorexia by using various interventions, but no these reportedly had consistent impact on the nutritional state.6, 7 Therefore, there is considerable interest in a possible role of ghrelin as an orexigenic peptide for the treatment of anorexia.21-24 Ghrelin is synthesized mainly in the upper third of the stomach with direct and indirect effects in the area postrema, the nucleus of the solitary tract, the nucleus of hypothalamus, and the brainstem, but it also may have direct and indirect metabolic effects on cells outside the central nervous system.25-32 Accordingly, providing ghrelin to tumor-bearing animals has produced evidence that ghrelin promotes appetite in cancer and other wasting conditions.33-35 Moreover, providing ghrelin as an intravenous infusion to normal individuals and to patients with cancer also has produced evidence that ghrelin stimulates appetite acutely without side effects.12 Therefore, the current study was planned as a follow-up of previous experiments on animal provision and short-term infusions to patients with the objective of evaluating whether daily, long-term ghrelin may have beneficial clinical effects in unselected, weight-losing cancer patients. A clear limitation of our study design was that ghrelin could be provided only once daily because of logistic and financial reasons among outpatients. Thus, 31 individuals represent a comparatively limited number of patients for statistical power. Nevertheless, the study was regarded worthwhile as a first step toward producing eventual proof of concept for ghrelin and to gain experience from this kind of treatment. In addition, it is important to emphasize that applied statistical evaluations in a 2-group comparison over time represents a safe but comparatively weak model for demonstrating any significant effects in 31 patients. All measurements were evaluated initially with an intention to treat; however, outcome results are presented per protocol treatment for those patients who were able to complete 8 weeks of treatment.
All of our invited and participating patients belonged to a group of weight-losing patients with subjectively severe anorexia for which no specific tumor therapy was regarded meaningful or effective according to our previous studies.1-3, 5 In addition, our patients could not have diabetes and could not be receiving steroid treatment or any other pharmacologic or nutritional support that had any importance with regard to outcomes in the current study. In fact, these kinds of patients are comparatively rare, even in a large surgical and oncologic institution. Thus, it is obvious that our study patients (high-dose group) and our control patients (low-dose group) may represent heterogeneous groups, although we stratified their inclusion by a using computerized randomization algorithm to minimize the risk of severe heterogeneity in important variables related to appetite, metabolic rate, nutritional state, and biochemical tests.20 Despite this precaution, it is possible that the patients who received high-dose ghrelin differed significantly from control patients (low-dose ghrelin) in some important respects. The worry may be that study patients seemed to suffer less from depression and anxiety, as evaluated by using standardized items at inclusion (Table 3), and that the history of weight loss was 5% lower among patients in the high-dose group, although the BMI was not different. It is also possible that a dose of 13 μg/kg daily of ghrelin actually was too low for subcutaneous injections compared with the doses provided by others reported in the literature from studies on healthy volunteers and intravenous infusions to cancer patients.12 However, the most important limitation in our design probably was the infrequent daily provision of ghrelin for the support of long-term improvements in patients with progressive disease.
With all of these limitations in mind, it was evident that long-term provision of ghrelin did not cause any obvious systemic or local side effects in any patient. In addition, there was no indication that ghrelin injections caused any consistent or statistically significant alteration in tumor markers over time in patients who received high-dose ghrelin compared with patients who received low-dose ghrelin, in agreement with our similar observations during insulin treatment to cancer patients.3 It is important to emphasize this finding, because immune-histology and cell culture experiments have provided some evidence that ghrelin may stimulate tumor cell proliferation and invasiveness.36-38 It also has been reported that malignant tumors may produce metabolically active ghrelin in tumor cells that express ghrelin receptors, suggesting the presence of paracrine loops, which are observed particularly in endocrine tumors.36 In support of such observations, 2 of our control patients (low-dose group) had surprisingly and unfortunately very high levels of endogenous ghrelin.39-42 However, those patients were not identified as clinical outliers at trial initiation or during follow-up. Therefore, it remains uncertain whether their circulating ghrelin levels had normal biologic effects.43
A most consistent finding in this study was that appetite scores, whole-body fat retention, and serum GH levels improved significantly during 8 weeks of treatment in patients who received high-dose ghrelin compared with patients who received low-dose ghrelin when analyzed both on an intent-to-treat basis and according to the protocol. These findings occurred without confirmation of significantly improved food intake, a measurement that has comparatively high variation (5%-15%) in repeated standardized measurements. Thus, dual x-ray absorpitometry measurements of body composition probably are more sensitive and reproducible.1-3 Moreover, it is not certain that whole-body fat retention depends preferentially on increased food intake after ghrelin administration. It may depend on the direct effects of ghrelin on metabolism in regional fat depots31, 44, 45 or on other, unrecognized effects on cachexia. Estimations of calculated resting energy balance also supported a trend toward significantly improved energy balance over 8 weeks of treatment in patients who received high-dose ghrelin (P < .07), and the patients who received high-dose ghrelin were in energy balance (P < .05). Thus, expected findings from ghrelin treatment and subsequently altered GH levels, subjective appetite scoring, and evidence of improved energy retention for fat deposition and perhaps for fat-free, lean tissue are encouraging observations indicating that the long-term provision of ghrelin may improve energy status and metabolism in weight-losing cancer patients in future clinical applications. Previous investigations in cancer patients have confirmed increased peak levels of GH after intravenous infusions of ghrelin.11 Our current results on altered fasting and basal serum levels of GH suggest that counter-regulatory effects by ghrelin were preserved more by high-dose treatment compared with low-dose treatment, because increased whole-body catabolism usually is associated with increasing basal levels of serum GH. Improvements by ghrelin did not induce increased physical activity and, thus, were used preferentially for the protection of energy stores. However, there is a need for the development of more easily provided dosing methods before oral food intake considering the current practical difficulties with stimulating the appetite using ghrelin injections. Such a solution would be ghrelin provision as a nose spray. In conclusion, to our knowledge, this is the first study in cancer patients suggesting that the daily, long-term provision of ghrelin to weight-losing patients with solid tumors supports host metabolism and a sensation of improved appetite, which appears to improve energy status in cachectic cancer patients.
Acknowledgements
We acknowledge the primary healthcare providers and all home care teams in Gothenburg city for their kind support and care of all participating patients.
CONFLICT OF INTEREST DISCLOSURES
Supported in parts by grants from the Swedish Cancer Society (2014), the Swedish Research Council (08712), Assar Gabrielsson Foundation (AB Volvo), Jubileumskliniken Foundation, IngaBritt and Arne Lundberg Research Foundation, Swedish and Gothenburg Medical Societies and the Medical Faculty, University of Gothenburg, Sahlgrenska University Hospital.
