The effects of vapreotide, a somatostatin analogue, on gastric acidity, gallbladder emptying and hormone release after 1 week of continuous subcutaneous infusion in normal subjects

Abstract

Aims

Somatostatin analogues (e.g. vapreotide) are used for treatment of acromegaly, endocrine tumours and variceal bleeding. The pharmacodynamic effects of vapreotide have, however, not been documented in the gastrointestinal tract. The aim of this study was to investigate the effects of continuous vapreotide administration on gastric acidity, gallbladder contraction and hormone release.

Methods

Ten healthy males participated in this randomised, placebo-controlled, double-blind, crossover trial. A constant vapreotide (or placebo) infusion (1.5 mg day⁻¹ s.c.) was given for 7 days with a portable pump. Intragastric pH was monitored on days 2 and 7. Gallbladder volume was sonographically assessed and the maximal ejection fraction was calculated. In addition basal and postprandial plasma levels of gastrin and cholecystokinin (CCK) were measured.

Results

After an initial increase in the median 24 h intragastric pH to a value of 2.6 on day 2, vapreotide’s effect on pH decreased: (day 7: median pH=1.9; respective placebo values were 1.7 and 1.5). On the same days with vapreotide treatment, gallbladder contraction and plasma levels of CCK were reduced; maximal ejection fractions after meal stimulation were 18% and 20% (respective placebo values were 57% and 62%). Plasma gastrin levels were not changed with vapreotide treatment.

Conclusions

The short lasting effect of vapreotide on intragastric acidity suggests a down-regulation of somatostatin receptors during treatment. The lack of effect on gastrin indicates that the effects on gastric pH are not mediated by gastrin. Constant vapreotide infusion (but not placebo) reduced gallbladder contraction suggesting a long-lasting effect on biliary function.

Introduction

The inhibitory effects of natural and synthetic somatostatin analogues on pituitary functions and on the gastrointestinal tract have been well documented in the last decade [1]. These compounds are used for the treatment of acromegaly [2], various endocrine tumours [3] and upper gastrointestinal bleedings [4]. Short-term systemic application of somatostatin has inhibitory effects on gastric acid secretion, gallbladder contraction and other gastrointestinal functions [5]. One main problem in chronic treatment with natural somatostatin is its short half-life and the need for intravenous application. The development of synthetic analogues has solved some of these problems. Vapreotide is an octapeptide analogue of natural somatostatin with a half-life of 164±46 min after a single subcutaneous injection (N. Lahlou, unpublished results). Repetitive subcutaneous injections three times daily are therefore needed for long-term treatment, a method poorly tolerated by patients and, with repetitive plasma peaks, not imitating a physiological hormonal state. However, during long-term treatment with octreotide, another somatostatin analogue, it has been noted that certain inhibitory effects are reduced after repeated application [6] suggesting a down-regulation of receptors. A systematic evaluation of the effect on different gastrointestinal organ systems in the same subject at stable plasma concentrations has, however, not yet been carried out. Therefore, throughout the course of 1 week, we followed the effect of vapreotide during chronic treatment on intragastric pH and gastrin, the main regulatory hormone of gastric acid production. Parallel to this, we also evaluated the effect of vapreotide on postprandial gallbladder contraction during chronic treatment in order to indirectly estimate the risk of promoting gallstone formation. Since vapreotide could play this role by inhibiting cholecystokinin (CCK), the main regulating hormone of meal-induced gallbladder contraction [7], we simultaneously measured it during the study. To obtain stable plasma concentrations, vapreotide was given as a continuous subcutaneous infusion using a small portable pump. Repetitive plasma concentrations of vapreotide were measured during the treatment periods to control the stable function of the pump. Growth hormone and IGF-1 were also measured to assess the known pharmacological activity of vapreotide.

Methods

Subjects

Ten healthy male volunteers, with a median age of 25 years (range 23–39), having a median body weight of 83 kg (range 64–93) and a median height of 183 cm (range 178–197), were selected for this trial. None of the subjects had any history of previous gastrointestinal, metabolic or endocrine disorders, and each had a normal screening physical examination and normal laboratory blood tests including serum chemistry, complete blood count and an electrocardiogram. All subjects were H. pylori negative based on a negative ¹³C-urea breath test and no gallbladder pathology including gallstones was found by abdominal ultrasound at the screening visit. The study was carried out in accordance with the recommendations of the Declaration of Helsinki [8] and was approved by the Ethics Committee for Human Research of the University Hospital of Basel. Written informed consent was obtained from each subject.

Experimental design

The study was carried out as a randomised, double-blind, cross-over trial. In brief, each subject had to take part in two treatment periods with continuous subcutaneous infusion of either placebo or vapreotide (1.5 mg day⁻¹) over 7 days. The two periods were separated by a wash-out period of at least 1 week (range 7 to 23 days).

To achieve constant drug plasma levels, the drug was infused subcutaneously with a Chronoclip® pump, a new infusion pump that allows delivery of 1 ml day⁻¹ for 1 week from a reservoir. This administration maintains plateau concentrations of about 2000 pg ml⁻¹ throughout the entire infusion period [9].

On day 1 after overnight fasting, the subcutaneous infusion pump was fixed by use of an infusion set-up to the left upper quadrant of the abdomen.

On day 2 and day 7 of each treatment period, 24 h intragastric pH, gallbladder contraction and plasma concentration of hormones were measured. To assess intragastric pH, a miniature combined glass electrode for continuous monitoring of intragastric pH (Ingold, CH-8920 Urdorf, Switzerland) was inserted transnasally in the fundus-corpus area of the stomach after calibration at 20° C using a commercial buffer solution, and temperature correction to 37° C. The electrode was connected to a portable data-recorder (MIC, CH-4500 Solothurn, Switzerland) with a sampling rate of one sample every 6 s. After overnight fasting, pH monitoring started at 08.30 h. Standardised meals were given after correct placement of the electrode (breakfast), at 14.30 h (500 ml of a liquid meal, Ensure®) and at 20.00 h (supper). The subjects were allowed to drink additional water.

Gallbladder volume was assessed by high resolution real-time sonography using a 3.5 MHz sector scanner (Aloka SSD650). The probe was first placed horizontally in the right upper quadrant of the abdomen while the subject was in supine position, then the probe was rotated to obtain the maximal possible diameter of the gallbladder on the screen. The machine’s software directly calculated the gallbladder volume by the sum of the cylinder method, previously described by Everson [10]. In the basal period, as well as after liquid meal ingestion, three measurements were performed at each time point (0,15, 30, 60, 90 and 120 min after 500 ml Ensure®) and the mean volume was taken for further calculations. At the same time, when gallbladder sonography was performed, separate venous blood samples were taken for determination of plasma gastrin and cholecystokinin. Venous blood samples for GH were drawn daily in the morning during both periods of treatment. Additional IGF-1 was measured before and after each treatment period.

Plasma concentrations of gastrin and CCK were determined by radioimmunoassay, as previously described [11]. The detection limit of the assays was 25 pg ml⁻¹ for gastrin and 3 pmol l⁻¹ for CCK. Plasma concentrations of GH and IGF-1 were determined by specific radioimmunoassays with a detection limit of 0.05 mU l⁻¹ for GH and 0.8 μg l⁻¹ for IGF-1.

Before installing the infusion pump on day 1, as well as at 6 and 12 h later, and, in addition, in the morning of every following day, venous blood samples were drawn for determination of vapreotide plasma levels. Additional samples were taken after disconnecting the pump until 48 h after stopping the infusion. The plasma concentration of vapreotide was measured using a radioimmunoassay method adapted from Mason-Garcia [12].

Materials

Vapreotide acetate was provided by Debiopharm (CH-1000 Lausanne 9, Switzerland); Ensure® (a standard liquid test meal with 16.7% protein, 30.1% lipid and 53.2% carbohydrates) was purchased from Abbott AG (CH-6330 Cham, Switzerland). Chronoclip®, a new portable infusion pump (40 g when filled) which can deliver small amounts of a drug (1 ml day⁻¹), was provided by Debiotech, CH-1000 Lausanne 9, Switzerland. For each experiment, the reservoir was filled with 10 ml of sterile infusion solution and connected through the motor unit to a set-up for subcutaneous infusion (Soft Set®).

Calculations and statistical analysis

Intragastric pH measurements were recorded every 6 s over 24 h. Median pH value was calculated for each subject as previously described [13]. Distribution of these values was characterised for each treatment group by median pH as well as 5th, 25th, 75th and 95th percentiles and the data are presented as box Whisker plots.

Basal gallbladder volumes in ml were compared using Student’s t-test. For further calculations, basal gallbladder volumes were taken as 100% and postprandial changes in gallbladder volume are given as % changes (mean± s.e.mean) from basal volume. Plasma concentrations of gastrin, CCK, GH and IGF-1 are given as mean± s.e.mean.

For statistical analysis, the area under the curve (AUC) of each hormone parameter as well as the maximum and the steady state plasma concentration of vapreotide were calculated for each individual subject and experiment. For comparison of parameters between vapreotide and placebo treatment and, on the other hand, between the first and last treatment day within a treatment group, data were first tested for normal distribution by the Wilk-Shapiro test. When normal distribution could not be rejected, data were analysed by analysis of variance (ANOVA), otherwise ANOVA was applied on rank-transformed data. In cases of significant differences, ANOVA was followed by the Duncan multi-comparison test for pair-wise comparisons. For all statistical analysis, SPSS for Windows Software was used. The level of statistical significance was P=0.05.

Results

Intragastric acidity

Median intragastric pH profiles over 24 h obtained during placebo and vapreotide treatment are shown in Figure 1. The pH-distribution-curve of vapreotide and placebo is shown in Figure 2; additional to the median value (50th percentile), the 5th, 25th, 75th and 95th percentile of 24 h pH-metry are shown in Whisker box plots (Figure 3). As can be seen, with placebo similar results were obtained on both days (median pH over 24 h was 1.7 on day 2 and 1.5 on day 7, NS). During the daytime, postprandial buffer effects were seen and during the night, typical individual variability was measured with a median pH slightly higher than during daytime.

Figure 1.

Median 24 h pH-profiles during subcutaneous infusion of placebo (a) or vapreotide (b) in 10 healthy volunteers.

Figure 2.

pH-distribution-curve of vapreotide (— day 2, –– day 7) and placebo (day 2, ––– day 7) in 10 healthy volunteers.

Figure 3.

Whisker box plots of 24 h pH-profiles after subcutaneous infusion of vapreotide or placebo in 10 healthy volunteers.

However, median pH values on both day 2 and day 7 were significantly (P < 0.01) higher in the vapreotide group than those measured during placebo administration (day 2: median 2.6 and day 7: median 1.9; Figure 1 and Whisker box plots in Figure 3).

On day 7, however, the median pH in the vapreotide group was significantly lower than on day 2 (P < 0.01). The dose of vapreotide used in this study did not raise the median intragastric pH above 4 in any of the 10 subjects (Figure 2).

Gallbladder volume

In the control period during placebo administration, basal gallbladder volumes were comparable on both days: 17±2 ml on day 2, respectively 15±2 ml on day 7 (Ratio 1.2; 95% CI=0.9–1.5). No significant differences were found postprandially with placebo treatment between day 2 and day 7. After ingestion of 500 ml of a liquid meal, gallbladder volume reached its minimal residual volume in the placebo group within 60 min, as can be seen in Figure 4. The ejection fraction at the time of maximal contraction was 63±10% on day 2 and 64±9% on day 7, respectively (NS).

Figure 4.

Postprandial gallbladder volume on day 2 (□) and 7 (•) during 1 week treatment with subcutaneous infusion of either a) placebo or b) vapreotide (1.5 mg day⁻¹) in 10 healthy volunteers. Data given as percent change of basal volume (mean±s.e.mean).

During vapreotide treatment, there was a significant enlargement of basal gallbladder volume to 25±3 ml (day 2, P < 0.05 vs placebo) and 29±2 ml (day 7, P < 0.01 vs placebo). There was no significant difference between the two latter volumes. A significant reduction of the residual volume 60 min after the meal, compared with basal gallbladder volume (21±3 ml on day 2, respectively 24±3 ml on day 7) was observed in the period with vapreotide treatment, but this reduction was much smaller than that observed during placebo treatment. Both values during vapreotide infusion on day 2 and day 7 were significantly (P < 0.01) higher than each of the corresponding values in the placebo group. The maximal ejection fraction was similar on day 2 and on day 7 (18±4% and 20±6%, respectively). During both treatment periods, gallbladder volume did not reach normal basal values within 2 h after the meal (Figure 4).

Hormone secretion

Plasma hormone levels of gastrin and CCK were measured parallel to the postprandial intragastric pH and gallbladder contraction after a liquid lunch meal. There was no difference in postprandial gastrin secretion in both groups (Table 1).

Table 1.

Postprandial gastrin and cholecystokinin (CCK) secretion during vapreotide or placebo infusion in 10 healthy volunteers (data are given as AUC (0, 120 min), mean±s.e.mean) and ratio with 95% confidence intervals (CI).

The fasting baseline concentrations of gastrin did not differ in the placebo group comparing day 2 and day 7 (35±2 pg ml⁻¹, 33±2 respectively). In the group treated with vapreotide, however, the basal concentrations of gastrin were significantly higher on day 2 (40±3 pg ml⁻¹; P < 0.05 vs placebo and P < 0.01 vs day 7) but no longer differed from the placebo group on day 7 (34±2 in the vapreotide group).

Fasting secretion of CCK did not differ between the treatments or within each treatment between day 2 and day 7.

Postprandial secretion of cholecystokinin was, however, slightly reduced during treatment with vapreotide (Table 1). While this difference was not significant, a similar trend could be shown on both days (Figure 5).

Figure 5.

Mean (±s.e.mean) postprandial cholecystokinin concentration (pmol l⁻¹) on day 2 (□) and day 7 (•) during subcutaneous infusion of a) placebo or b) vapreotide in 10 healthy volunteers.

In the control period during placebo administration, the AUC of plasma concentration of GH was 112±163, while in the period during vapreotide administration it was 52±55 (P < 0.01 vs placebo).

Mean baseline concentrations of IGF-1 before the administration were similar in both treatment groups: 231±73 μg l⁻¹ and 244±58 μg l⁻¹ in the placebo and vapreotide group, respectively. During placebo treatment there was no change in the IGF-1 concentrations on day 7 (249±75 μg l⁻¹) as compared with values before administration. In contrast to this, IGF-1 concentrations decreased significantly (P < 0.01) to 178±46 μg l⁻¹ during subcutaneous administration of vapreotide. The latter value was also significantly (P < 0.01) lower than IGF-1 concentration on day 7 in the placebo group.

Vapreotide plasma levels

During the 7 days of treatment with a constant subcutaneous infusion of 1.5 mg day⁻¹, steady state plasma concentration of 2062±180 pg ml⁻¹ could be achieved. Maximal plasma concentration was 3366±527 pg ml⁻¹, the area under the curve 406±35 ng ml⁻¹ h.

Side effects

Some subjects complained of steatorrhoea and abdominal pain of a mild to moderate degree, but only one of them had to be treated with a spasmolytic drug for abdominal cramps, most probably due to an irritable colon (same complaints in verum and placebo phases).

Discussion

Intragastric pH was markedly elevated by vapreotide, but only at the beginning of a long term treatment, while the elevation was much lower after 1 week. The efficacy concerning gallbladder contraction, however, did not disappear after 1 week of treatment with vapreotide. Basal gastrin concentrations also changed only at the beginning of the vapreotide phase, being significantly higher than in the placebo group on day 2, while there was no change after 1 week of treatment. Postprandial gastrin levels were not affected by the treatments. Postprandial cholecystokinin levels were slightly reduced at the beginning of the vapreotide phase and after 1 week of treatment. Using a pump for continuous subcutaneous administration of vapreotide, the plasma levels of the drug were maintained at about 2000 pg ml⁻¹ throughout the entire period.

The present study has shown a marked elevation of intragastric pH by the synthetic somatostatin analogue vapreotide. This effect could be demonstrated on day 2, but only to a much smaller extent on day 7 after continuous vapreotide infusion during 1 week, suggesting tachyphylaxis. A similar disappearance of the efficacy, with regard to intragastric pH after chronic treatment, has been previously demonstrated with octreotide, another somatostatin analogue, given as rapid s.c. injection twice daily by Londong and coworkers [6], but the present study was the first using continuous infusion. The reason for the loss of efficacy after several days of treatment with a synthetic somatostatin analogue remains unclear.

Since plasma gastrin levels were not reduced under treatment conditions, it seems that the inhibitory effect of vapreotide on intragastric pH is not mediated by suppression of gastrin, the main regulatory peptide of gastric acid secretion. This suggests a direct inhibition of histamine producing cells and/or parietal cells in the stomach, and this inhibition shows a tachyphylaxis.

The initial rise in basal gastrin levels on day 2 in the vapreotide group is explained by the lack of negative feedback due to the elevated intragastric pH. On day 7, however, when intragastric pH in both groups was almost similar, plasma gastrin levels no longer differed. Postprandial plasma gastrin levels were not affected by the treatment since they are directly stimulated by intragastric calcium and amino acids.

A second aim of the present study was to investigate the behaviour of the gallbladder during chronic treatment with vapreotide. In contrast to the effect on intragastric pH, no tachyphylaxis occurs and the contraction of the gallbladder is equal on day 2 and on day 7 of the treatment. These data are in contrast to findings of the group who also showed an enlargement of basal gallbladder volume at the beginning and at the end of 1 week of octreotide treatment twice daily, but no further difference in kinetics of gallbladder contraction after stimulation with exogenously administrated CCK [14].

These differences can best be explained by the different kind of stimulation: while Creutzfeldt used exogenous CCK, we measured gallbladder contraction after a meal which acts as a physiological stimulant. In our experiments, plasma CCK levels were inhibited postprandially in the vapreotide group, but these results were not significant. This is supported by additional experiments reported by the same group [14] after 1 week of pre-treatment with octreotide, when the stimulant for gallbladder contraction was a meal. These data also show a marked inhibition of the contractile biliary response and, at the same time, an abolished CCK secretion. Therefore, CCK plays a major role in postprandial gallbladder contraction, and its inhibition, even if not significant in the present study, is important in regulating gallbladder contraction by somatostatin analogues.

In summary, it can be concluded that gallbladder stasis occurs not only at the beginning of treatment with somatostatin analogues, but also during prolonged use of a constant amount. Gallbladder stasis is a high-risk factor for gallstone formation and provides important clinical implications, not only for the chronic treatment of oesophageal varices as a potential indication for vapreotide, but also for its therapeutic use in endocrine active tumours. We could also demonstrate that with the Chronoclip® pump, a subcutaneous infusion of very small drug amounts can be constantly infused 24 h a day, and that sufficient and stable plasma levels of vapreotide can be produced throughout 1 week. These plasma levels were able to suppress GH during the entire period of vapreotide infusion showing a pharmacological activity of the drug. No major side effects appeared in the verum period and the system was well tolerated by the volunteers.