Influence of polymers ratio on insulin-loaded nanoparticles based on poly-epsilon-caprolactone and Eudragit RS for oral administration.

Nanoparticles loaded with two different commercial insulins (Actrapid, Novorapid and based on different blends of a biodegradable polyester (poly-epsilon-caprolactone) and a polycationic non-biodegradable acrylic polymer (Eudragit RS) were characterized in vitro. The zeta potential was positive whenever Eudragit RS was part of the nanoparticles matrix. The encapsulation efficiency was ~ 96% except for Novorapid-loaded particles of poly-epsilon-caprolactone (only 35%). In vitro release studies revealed a burst release from nanoparticles, which may be of interest for oral delivery. Novorapid-loaded nanoparticles were orally administered to diabetic rats and allowed the glycemia to be decreased when compared with free nanoparticles.

Stealth nanoparticles coated with heparin as peptide or protein carriers.

Nanoparticles (prepared from a mixture of polyester and a polycationic polymer) loaded with insulin were prepared by a double emulsion method followed by evaporation solvent. Low molecular weight heparin (LMWH) was bound by electrostatic interactions onto the surface of the particles to confer Stealth properties. These nanoparticles were characterized in vitro (mean diameter, zeta potential, encapsulation efficiency, and release kinetics) and compared with conventional (without LMWH) and unloaded nanoparticles. The pharmacokinetics of insulin were studied after intravenous injection into diabetic rats in the form of Stealth or conventional nanoparticles or as a solution. Stealth nanoparticles allowed an increase in the elimination half-life of insulin, showing that the hydrophilic layer of LMWH was able to limit recognition by the mononuclear phagocytosis system in vivo. However, complement activation studies (CH50) did not reveal significant difference between Stealth and conventional nanoparticles.

Formulation of insulin-loaded polymeric nanoparticles using response surface methodology.

The objective of this work was to formulate new oral insulin-loaded nanoparticules using the response surface methodology. The insulin nanoparticles were prepared by a water-in-oil-in-water emulsification and evaporation method. The polymers used for the encapsulation were blends of biodegradable poly-epsilon-caprolactone (PCL) and of positively-charged, nonbiodegradable polymer (Eudragis RS). A central composite design has been built to investigate the effects of three controlled variables: ratio of polymers (PCL/RS ratio), volume, and pH of the aqueous solution of polyvinyl alcohol. The nanoparticles were characterized by measuring the amount of entrapped insulin, the particle size, the polydispersity of the obtained particles, the zeta potential, and the amount of insulin released after 7 hours. A second-order model was evaluated by multiple regression and was statistically tested for each of the studied controlled variable. The obtained polynomials proved efficient to localize an optimal operating area highlighted by the use of three-dimensional response surfaces and their corresponding isoresponse curves. An interesting formulation given by the models was selected, prepared, and evaluated. The corresponding quantity of entrapped insulin was 25 IU per 100 mg of polymer, and the particle size was 350 nm with a polydispersity of 0.21. The quantity of released insulin was 4.8 IU per 100 mg of polymer after 7 hours and the zeta potential was +44 mV. All these collected values were in perfect accordance with values estimated by the models. Finally, the results suggested that PCL/RS 50/50 nanoparticles might represent a promising formulation for oral delivery of insulin.

Ileal uptake of polyalkylcyanoacrylate nanocapsules in the rat.

The ileal uptake of polyalkylcyanoacrylate nanocapsules (less than 300 nm in diameter) has been investigated in the rat. Iodised oil (Lipiodol) was used as the tracer for X-ray microprobe analysis in scanning electron microscopy. Lipiodol nanocapsules, or an emulsion of Lipiodol, were administered in the lumen of an isolated ileal loop of rat. Lipiodol nanocapsules improved the absorption of the tracer as indicated by increased concentrations of iodine in the mesenteric blood (+27%, P < 0-01, compared with Lipiodol emulsion). Intestinal biopsies were taken at different time points and the samples underwent cryofixation and freeze-drying. The nanocapsules were characterized by their strong iodine emission, and electron microscopy of the biopsy samples revealed nanocapsules in the intraluminal mucus of the non-follicular epithelium, then in the intercellular spaces between enterocytes, and finally the nanocapsules were found within intravillus capillaries. However, nanocapsules were most abundant in the Peyer's patches, where the intestinal epithelium had been crossed by way of the specialized epithelial cells, designated membranous cells, or M cells, and their adjacent absorptive cells. These observations were confirmed quantitatively by measuring iodine concentrations in the various tissue compartments. Ten minutes after the intraluminal administration of Lipiodol nanocapsules, the emission of iodine peaked in the mucus (+77%, P < 0.01), in M cells (+366%, P <0.001), in enterocytes adjacent to M cells (+70%, P < 0.05) and in lymph vessels (+59%, P < 0.05). Polyalkylcyanoacrylate nanocapsules were able to pass through the ileal mucosa of the rat via a paracellular pathway in the non-follicular epithelium, and most predominantly, via M cells and adjacent enterocytes in Peyer's patches.

Effects of cholecystokinin octapeptide on a pancreatic acinar carcinoma in the rat.

PURPOSE: To investigate the effects of increasing concentrations of cholecystokinin octapeptide (CCK-8) on a pancreatic acinar adenocarcinoma. METHODS: Growth of the tumour was estimated in vivo on rats bearing a subcutaneous pancreatic carcinoma, and in vitro on primary cultured tumour cells. CCK receptors were characterized by binding assays. RESULTS: CCK-8, administered for 12 successive days, exerted a biphasic action on tumour growth: a dose-dependent stimulation with low doses (0.1 and 0.5 microg/kg) and inhibition with high doses (2 and 4 microg/ kg) as shown by respective increases and decreases in tumor volume, protein, RNA and amylase contents. In cell cultures, [3H]thymidine incorporation was dose-dependently increased with 10-(10) to 10(-8) M CCK-8 and inhibited with 10(-7) M. Both effects were completely suppressed by the CCK-receptor antagonists CR 1409 and L 364,718 (10(-4) M). Binding studies showed the overexpression of two classes of CCK-A receptors of low and high affinity when compared to the normal pancreas which was less sensitive to CCK-8. CONCLUSIONS: CCK-8 exerts a biphasic growth response on the acinar pancreatic carcinoma, mediated by two classes of CCK-A receptors overexpressed in the tumour.

Effect of the gastrin-releasing peptide antagonist BIM 26226 and lanreotide on an acinar pancreatic carcinoma.

The effects of a potent specific gastrin-releasing peptide receptor antagonist, BIM 26226 ([D-F5 Phe6, D-Ala11] bombesin (6-13) OMe), and the long-acting somatostatin analogue, lanreotide (BIM 23014), on the growth of an acinar pancreatic adenocarcinoma growing in the rat or cultured in vitro were investigated. Lewis rats bearing a pancreatic carcinoma transplanted s.c. in the scapular region, were treated with gastrin-releasing peptide (30 microg/kg per day), BIM 26226 (30 and 100 microg/kg per day) and lanreotide (100 microg/kg per day) alone or in combination for 14 successive days. Chronic administration of BIM 26226 and lanreotide significantly inhibited the growth of pancreatic tumours stimulated or not by gastrin-releasing peptide (GRP), as shown by a reduction in tumour volume, protein, ribonucleic acid, amylase and chymotrypsin contents. This effect was more pronounced with 100 microg/kg per day BIM 26226 than with 30 microg/kg per day. However, BIM 26226 and lanreotide, given together, did not exert any additive effect on GRP-treated and -untreated tumours. In cell cultures, both BIM 26226 and lanreotide (10(-6) M) inhibited [3H]thymidine incorporation in tumour cells induced or not by GRP, but no increased effect was observed after combined treatment with both agents. Binding studies showed that BIM 26226 had a high affinity for GRP receptors in tumour cell membranes (IC50 = 6 nM). These results from in vivo and in vitro experiments suggest that BIM 26226 and lanreotide are able to reduce the growth of an experimental acinar pancreatic tumour. Thus, these agents represent interesting steps toward the development of new approaches for treatment of pancreatic carcinomas.

Poly(alkyl cyanoacrylate) nanocapsules as a delivery system in the rat for octreotide, a long-acting somatostatin analogue.

Poly(alkyl cyanoacrylate) nanocapsules have been used as biodegradable polymeric drug carriers for subcutaneous and peroral delivery of octreotide, a long-acting somatostatin analogue; their ability to reduce insulin secretion or prolactin secretion in response to oestrogens has been studied in adult male rats. The nanocapsules, prepared by interfacial emulsion polymerization of isobutyl cyanoacrylate, were 260 nm in diameter and incorporated 60% of octreotide. Administered subcutaneously, the octreotide-loaded (20 micrograms kg-1) nanocapsules suppressed the insulinaemia peak induced by intravenous glucose overload and depressed insulin secretion over 48 h, preventing the secretory rebound; however, glycaemia was unaffected. In parallel, the plasma octreotide concentration increased 2.7 times. Administered perorally to oestrogen-treated rats, octreotide-loaded nanocapsules (200 and 1000 micrograms kg-1) significantly improved the reduction of prolactin secretion (by 72 and 88%, respectively, compared with 32 and 54% with free octreotide) and slightly increased plasma octreotide level. Thus nanocapsules could be of interest as a biodegradable drug carrier for the administration of octreotide.

Poly(alkyl cyanoacrylate) nanospheres for oral administration of insulin.

Poly(alkyl cyanoacrylate) nanocapsules have been successfully used for oral administration of insulin in diabetic rats. This work reports a suitable formulation for insulin-loaded nanospheres composed of full polymeric structures formed by polymerization of isobutyl cyanoacrylate (IBCA) in an acidic medium, insulin (15 U/mL) being added to the polymerization medium 60 min after the onset of polymerization. These nanospheres (MW 364) displayed a mean size of 145 nm and an association rate of 1 U of insulin/mg of polymer. They protected insulin from the degradation by proteolytic enzymes in vitro, especially when they were dispersed in an oily medium (Miglyol 812) containing surfactive agents (Poloxamer 188 and deoxycholic acid). When dispersed in the same medium, insulin-loaded nanospheres (100 U/kg of body weight), administered perorally in streptozotocin-induced diabetic rats, provoked a 50% decrease of fasted glycemia from the second hour up to 10-13 days. This effect was shorter (2 days) or absent when nanospheres were dispersed in water with surfactive agents or not. Using 14C-labeled nanospheres loaded with [125I]insulin, it was found that nanospheres increased the uptake of [125I]insulin or its metabolites in the gastrointestinal tract, blood, and liver while the excretion was delayed when compared to [125I]insulin nonassociated to nanospheres; in addition, 14C- and 125I-radioactivities disappeared progressively as a function of time, parallel to the biological effect. Thus insulin-loaded nanospheres can be considered as a convenient delivery system for oral insulin at the prerequisite that they were dispersed in an oily phase containing surfactants.

Intestinal absorption of PLAGA microspheres in the rat.

Rhodamine B-labelled poly (DL-lactide-co-glycolide) (PLAGA) microspheres of 2 different sizes, 1-5 microns and 5-10 microns, were administered as a single dose (1.44 x 10(9) and 1.83 x 10(8) particles, respectively) into the ileal lumen of adult rats. The content of rhodamine in the mesenteric vein and ileal lumen was analysed periodically from 10 min to 48 h as well as the distribution of microspheres in the intestinal mucosa and various other tissues. The concentration of rhodamine decreased progressively in the intestinal lumen and was negligible after 24 h. The number of microspheres in the mesenteric vein increased rapidly and reached a maximum after 4 h whatever the size of the particles. It then decreased progressively, but more rapidly with microspheres > 5 microns than with microspheres < 5 microns. The absorption efficiency was low for the former batch (about 0.11% of the administered dose) and higher for the latter (about 12.7%). The intraileal administration of free rhodamine B was followed by intense labelling of the epithelial cells and basement membranes in mesenteric lymph nodes, spleen, kidney and liver. PLAGA microspheres mainly crossed the intestinal mucosa at the site of Peyer's patches where microspheres of < 5 microns appeared after 3 h. Microspheres > 5 microns were retained in the ileal lumen. A few small microspheres were occasionally observed in the epithelial cells. Only the smallest particles were recovered in the liver, lymph nodes and spleen while basement membranes were always labelled. It is concluded that PLAGA microspheres could be useful for the oral delivery of antigens if their size is between 1 and 5 microns.

Biomaterial supports for colonic wall defect healing. An experimental study in the rat.

New artificial biomaterials were tested for support of gastro-intestinal tract wound healing in the rat. Double layered collagenic matrices were prepared with purified collagens extracted from human placental tissues. Two types of patches were tested, the first constituted from a collagen type I + III layer covered by collagen IV in a liquid phase (patch I + III/IV) and the second from a collagen IV layer covered by liquid collagen IV (patch IV/IV). The matrices were applied with fibrin sealant to the edges of a 1 cm diameter colonic wall defect in the rat. Healing evolution was determined by macroscopic, microscopic and immunostaining studies. The reconstitution of the three colonic wall layers was achieved within 45 days without retraction or inflammatory reaction, while the biomaterial was resorbed. Human collagen I and III antibodies failed to stain extracellular matrix. This failure may be a consequence of outdated antibodies or more likely epitope alteration during extraction and preparation of the collagens. A human collagen type IV antibody staining of the scar zone showed the basement membranes of newly developed vessels within 10 days, and newly formed colonic mucosa within 20 days. The collagen reconstituted matrix was able to assist healing of normal digestive tract defects as shown by the labelling of the new synthesized extracellular matrix by collagen type IV antibody. These findings support the use of collagen biomaterial in gastro-intestinal anastomosis. This new surgical approach allowing healing of colonic wall defects could reduce occurrence of anastomotic leakage in human.

Biomaterials for primary closure of a choledochotomy in dogs.

Since primary closure of the common bile duct is often not undertaken because of the risks of biliary leakage and peritonitis, we have evaluated feasibility and reliability of closure using biomaterials. In three groups of dogs, an unsutured choledochotomy was closed with circular glued patches: a scleroprotein patch in 4 dogs and an oxidized, compressed human collagen patch reinforced (n = 6) or not (n = 6) with three stitches. The scleroprotein patch (n = 4) was resorbed too soon, and in 2 dogs the unstitched collagen patches became unglued; biliary leakage was the result in both instances. The bile duct healed successfully within 1 month in the other 10 animals fitted with collagen patches, despite one common bile duct stricture. Safe primary closure of a choledochotomy may be envisioned in humans if the duct suture is protected by this new collagen biomaterial.

Expression and characterization of gastrin-releasing peptide receptor in normal and cancerous pancreas.

The biochemical and pharmacological characteristics of specific binding sites for gastrin-releasing peptide (GRP) were investigated in normal exocrine pancreas and in an azaserine-induced pancreatic carcinoma in the rat, under similar experimental conditions. Cells from both types of tissues contained rapid, reversible, temperature-dependent, and highly specific binding sites for GRP. Scatchard analysis of equilibrium data obtained with normal and tumor plasma membranes indicated a single class of high-affinity sites (KD = 0.42 +/- 0.06 and 0.35 +/- 0.05 nM, respectively), but the number of GRP receptors was significantly different (Bmax = 31 +/- 4.5 and 189 +/- 20 fmol/mg protein, respectively). Binding of 125I-GRP1-27 was sensitive to GTP analogues, suggesting that the GRP receptor is functionally linked to a guanyl regulatory protein; however, the wheat germ agglutinin-agarose purified receptor had lost this G-protein activity. Cross-linking of 125I-GRP1-27 either to normal and neoplastic cells or to crude membranes, solubilized membrane proteins, and partially purified receptors revealed the presence of a specific MW 75-kDa polypeptide. N-Glycanase treatment reduced this apparent MW to about 45 kDa. Together, these data suggest that normal and tumor pancreatic cells contain a specific GRP receptor that is expressed more on malignant pancreatic tissues.

Influence of corona surface treatment on the properties of an artificial membrane used for Langerhans islets encapsulation: permeability and biocompatibility studies.

An artificial membrane (AN69 Hospal) suitable for pancreatic islets encapsulation was submitted to a physicochemical treatment (corona discharge) to improve its insulin permeability. This effect depends on the duration of the electrical discharge (expressed as the speed of a conveyor belt) and the distance between the electrodes and the membrane. Among the various treatments tested, the most efficient (distance of 5 cm and a speed of 2 cm s-1) produced a three-fold increase in insulin diffusion. This improvement persisted after a protein-coating test which mimics in vivo conditions. At 1 y after the peritoneal implantation, the corona-treated membrane remained biocompatible. Thus, corona discharge treatment may serve to optimize the properties of artificial membranes used for pancreatic islets encapsulation.

Effect of n-hexacosanol on insulin secretion in the rat.

n-Hexacosanol, a long-chain saturated fatty alcohol extracted from Hygrophyla erecta Hochr., has been recently shown to exert neurotrophic properties on central neurons and to stimulate phagocytosis in macrophages. The present work was designed to investigate the effects of hexacosanol on stimulated insulin secretion in vivo and in vitro. In anaesthetized rats, hexacosanol (2 mg/kg i.p.) induced a reduction of the insulin response to an intravenous glucose tolerance test (0.3 g/kg) with a consequent increase in hyperglycaemia. In vitro, in the isolated perfused pancreas, hexacosanol at the concentration of 10(-7) M clearly reduced the two phases of glucose-induced insulin secretion. At the higher concentration (10(-5) M), hexacosanol was no longer able to exert an inhibition of glucose-induced insulin release; surprisingly a stimulating effect occurred which was of the same magnitude as in control experiments with Tween alone, at the concentration used to dissolve hexacosanol. In isolated perifused islets, 22 mM glucose-stimulated insulin release was also inhibited by hexacosanol at the concentrations of 10(-9) M and 10(-7) M, but not at 10(-5) M. In contrast, insulin secretion induced by arginine (20 mM) was not affected by the different concentrations of hexacosanol. It is concluded that n-hexacosanol at 10(-9) M and 10(-7) M exerts an inhibitory effect on insulin secretion stimulated by glucose in vivo and in vitro in the rat, suggesting a direct effect on islets of Langerhans.

Experimental pancreatic cancer in the rat treated by photodynamic therapy.

Selective histological necrosis of experimental pancreatic carcinoma by photodynamic therapy (PDT) has been successful with haematoporphyrin derivatives and phthalocyanine as photosensitizers. This report describes the feasibility of PDT with pheophorbide A as the photosensitizer to treat azaserine-induced pancreatic rat carcinoma and analyses survival of the animals. An organ distribution study 24 h after pheophorbide A administration (9 mg/kg intravenously) gave a selectivity ratio of 13.5:1 between tumour and surrounding tissue. Light of 660 nm and 100 J/cm2 induced selective necrosis of the tumour. Six of nine rats were cured in 120 days whereas all 36 control animals died within 35 days (P < 0.01). The pancrease and hepatic pedicle were relatively unaffected by PDT, but the duodenum was injured.

Implication of cholecystokinin in pancreatic adaptation after biliopancreatic bypass in the rat.

Biliopancreatic bypass (BPB), a bariatric surgical procedure, leads to a malnutrition-induced general visceral atrophy except for the pancreas. This work investigates the implication of cholecystokinin (CCK) in the exocrine pancreatic adaptive process using a plasma CCK assay and the CCK receptor antagonist CR 1409. No significant reduction in weight and DNA content of the pancreas was noted 36 days after BPB, while a strong decrease in protein, enzymes and RNA contents indicating cellular hypotrophy became apparent. CR 1409 treatment strongly depressed pancreatic weight and its DNA content in BPB animals, suggesting an additional hypoplasia; however, the reduction in pancreatic enzyme content was not aggravated. BPB increased plasma CCK concentrations by 160%, unrelated to CR 1409 treatment. These results indicate that: (1) CCK is involved in the pancreatic adaptive response after BPB in rats, and (2) in the context of a protein malnutrition state, CCK dissociates its pancreatic growth and enzymatic effects, favouring the former.

Trophic and enzymatic adaptation of the intestine to biliopancreatic bypass in the rat.

Biliopancreatic bypass (BPB), the exclusion of a duodenojejunal loop from the digestive continuity, has been proposed as a bariatric procedure for treatment of morbid obesity. The present study in rats investigated the effect of this surgical procedure on the mucosae of the ileum directly anastomosed to the stomach, and of the jejunum irrigated only by biliopancreatic secretions. The proximal part of the ileum adapted by two-fold increases in its mucosal mass, total protein and RNA content; DNA content was four-fold higher than in sham-operated animals. There was a correlated increase of mucosal enzyme content, except for lactase. In the distal ileal mucosae, a slight, transient augmentation of mucosal mass, protein, DNA and RNA content was observed which tended to compensate for the shortening of the functional gut. No morphological changes were found in the excluded loop, probably because of an endoluminal stimulation by biliopancreatic secretions. Thus, in BPB, biliopancreatic secretions seem to exert trophic effects on the intestinal mucosa, but they are less potent than the endoluminal nutrition that restores the oral-aboral mucosal gradient.

Distribution of pheophorbide A in normal tissues and in an experimental pancreatic cancer in rats.

The in vivo administration and distribution of a potent new photosensitizer, pheophorbide A (PH-A), was investigated in rats. The spectral characteristics were determined. This hydrophobic compound was solubilized by an ethanol/phosphate-buffered saline (PBS) mixture (v/v) and sonicated immediately before i.v. administration. Tissue distribution and the affinity of PH-A for an acinar pancreatic tumor were determined in Lewis rats for up to 48 h after a single i.v. administration of 3 mg kg-1 body wt. Methanol-extracted PH-A was quantitatively determined by fluorescence spectrophotometry at 665.6 nm. The PH-A uptake pattern showed that the reticulo-endothelial system is the major target of PH-A, followed by the gut and then the lung and pancreas. PH-A concentrations in skin were very low. The presence of an enterohepatic cycle was suggested by the PH-A biliary output, intestinal uptake and blood concentrations. Tumor PH-A retention was longer than pancreatic retention. The ratio of tumoral to peri-tumoral pancreas PH-A was 6.7:1, 24 h after i.v. administration. With its similar tissue pattern, better absorption spectrum and lower skin toxicity, PH-A could be a more potent photosensitizer than hematoporphyrin derivatives.

Inhibition of growth of a transplanted rat pancreatic acinar carcinoma with CCK-8.

Gastrointestinal hormones and neuropeptides are known to regulate growth of various normal gastrointestinal tissues and many cancers. Since cholecystokinin (CCK) is considered the most potent trophic factor for the exocrine pancreas, we studied its effect on growth of an acinar cell tumor, initially induced by azaserine and transplanted to the rat, in comparison with the normal pancreas. When tumors became palpable, rats were treated three times daily for 12 or 14 days with CCK-8 or NaCl 0.9% (controls) alone or in combination with the CCK receptor antagonist CR1409 (10 mg/kg) administered subcutaneously twice daily. Then tumors and pancreata were analyzed for their size, composition, and CCK receptors. Tumor volume, weight, and protein content, RNA, DNA, and enzymes decreased after CCK-8 treatment in a dose-dependent manner, the maximal effect being observed with 4-micrograms/kg treatment. This inhibitory effect was partially suppressed by CR1409, which by itself also reduced tumor growth, but to a lesser degree. CCK-8 exerted a stimulating effect on growth of the normal pancreas with low doses (1 and 2 micrograms/kg) and an inhibitory effect or no effect with a higher dose (4 micrograms/kg). CR1409 prevented this latter effect, but did not affect by itself the normal pancreas. These findings suggest that CCK-8 inhibits growth of an acinar cell tumor grafted to the rat; this effect is mediated by the occupation of specific CCK receptors present in high density on these cells. In contrast, CCK-8 exerts a biphasic effect on the normal pancreas as a function of its dose.

Gastrin-releasing peptide: in vivo and in vitro growth effects on an acinar pancreatic carcinoma.

The mammalian gastrin-releasing-peptide (GRP) and its structural amphibian analogue, bombesin, are known to be trophic factors for the normal exocrine pancreas. This work investigates the possible role of GRP in the growth of an acinar pancreatic cancer transplanted to the rat and in primary tumor cell cultures. Moreover, this adenocarcinoma was tested for its content of specific bombesin/GRP receptors by using autoradiographic technics and in vitro binding assays with tumor cells. In Lewis rats bearing the pancreatic carcinoma transplanted s.c. in the scapular region, chronic administration of GRP at the dose 30 micrograms/kg/day for 15 successive days significantly increased the tumor volume, the final tumor weight, and amylase, protein, RNA and DNA contents. Autoradiographic studies showed that tumor tissue was GRP receptor positive with a high density. The biochemical characterization indicated that receptor positive tumor tissue had saturable and high affinity receptors with pharmacological specificity for GRP and its bioactive analogues. In primary tumor cell cultures, GRP increased the incorporation of [3H] thymidine in DNA in a dose- and time-dependent manner. There was a good correlation between the ability of GRP and its COOH terminal analogues to elicit DNA synthesis and their affinity for 125I-GRP binding sites. These results from in vivo and in vitro experiments demonstrated that GRP induces growth of pancreatic carcinoma by acting directly on specific membrane receptors present on the tumor cells.