Early diagnosis of pancreatic cancer: neutrophil gelatinase-associated lipocalin as a marker of pancreatic intraepithelial neoplasia.

Pancreatic cancer is a highly lethal malignancy with a dismal 5-year survival of less than 5%. The scarcity of early biomarkers has considerably hindered our ability to launch preventive measures for this malignancy in a timely manner. Neutrophil gelatinase-associated lipocalin (NGAL), a 24-kDa glycoprotein, was reported to be upregulated nearly 27-fold in pancreatic cancer cells compared to normal ductal cells in a microarray analysis. Given the need for biomarkers in the early diagnosis of pancreatic cancer, we investigated the expression of NGAL in tissues with the objective of examining if NGAL immunostaining could be used to identify foci of pancreatic intraepithelial neoplasia, premalignant lesions preceding invasive cancer. To examine a possible correlation between NGAL expression and the degree of differentiation, we also analysed NGAL levels in pancreatic cancer cell lines with varying grades of differentiation. Although NGAL expression was strongly upregulated in pancreatic cancer, and moderately in pancreatitis, only a weak expression could be detected in the healthy pancreas. The average composite score for adenocarcinoma (4.26+/-2.44) was significantly higher than that for the normal pancreas (1.0) or pancreatitis (1.0) (P<0.0001). Further, although both well- and moderately differentiated pancreatic cancer were positive for NGAL, poorly differentiated adenocarcinoma was uniformly negative. Importantly, NGAL expression was detected as early as the PanIN-1 stage, suggesting that it could be a marker of the earliest premalignant changes in the pancreas. Further, we examined NGAL levels in serum samples. Serum NGAL levels were above the cutoff for healthy individuals in 94% of pancreatic cancer and 62.5% each of acute and chronic pancreatitis samples. However, the difference between NGAL levels in pancreatitis and pancreatic cancer was not significant. A ROC curve analysis revealed that ELISA for NGAL is fairly accurate in distinguishing pancreatic cancer from non-cancer cases (area under curve=0.75). In conclusion, NGAL is highly expressed in early dysplastic lesions in the pancreas, suggesting a possible role as an early diagnostic marker for pancreatic cancer. Further, serum NGAL measurement could be investigated as a possible biomarker in pancreatitis and pancreatic adenocarcinoma.

Class III beta-tubulin, a marker of resistance to paclitaxel, is overexpressed in pancreatic ductal adenocarcinoma and intraepithelial neoplasia.

AIMS: Class III beta-tubulin (TUBB3) reduces microtubule stability and confers resistance to microtubule-stabilizing taxanes, including paclitaxel and docetaxel. Pancreatic ductal adenocarcinomas show limited responsiveness to taxanes, but little is known of the underlying mechanisms. The aim of this study was to examine TUBB3 expression in pancreatic cancer cell lines, invasive pancreatic adenocarcinoma and pancreatic intraepithelial neoplasia (PanIN). METHODS AND RESULTS: Reverse transcriptase-polymerase chain reaction and Western blot were used to study TUBB3 expression in pancreatic cancer cell lines. Immunohistochemistry was employed to assess TUBB3 in pancreatic cancer specimens, including 75 invasive adenocarcinomas and 41 PanIN precursor lesions. TUBB3 was undetectable in non-neoplastic ducts of the pancreas. In contrast, the vast majority (78-93%) of pancreatic ductal adenocarcinomas demonstrated either diffuse or focal TUBB3 expression. TUBB3 was found to increase progressively in PanIN lesions from 3/16 of PanIN-1 (19%), 5/17 of PanIN-2 (29%) to 5/8 of PanIN-3 lesions (63%). CONCLUSIONS: TUBB3 is expressed in most pancreatic ductal adenocarcinomas, possibly accounting for the suboptimal response of these tumours to microtubule-stabilizing agents. Up-regulation of TUBB3 in PanIN lesions suggests that microtubule dysfunction is an early feature of this disease. TUBB3 immunohistochemistry could potentially help identify pancreatic cancer patients lacking TUBB3 expression who might benefit from taxane therapy.

Retinoic acid can induce markers of endocrine transdifferentiation in pancreatic ductal adenocarcinoma: preliminary observations from an in vitro cell line model.

BACKGROUND AND HYPOTHESIS: The pancreatic ductal adenocarcinoma (HPAF) cells have a multipotent stem cell potential. It was hypothesised that all-trans-retinoic acid (atRA) can induce transdifferentiation of these cells into cells with an endocrine phenotype. MATERIAL AND METHODS: To explore this hypothesis, an in vitro system of cells was established. Some cells were treated with atRA at concentrations of 100 nmol/l (non-apoptosis-inducing) and 5 micromol/l (apoptosis-inducing) and harvested. Cells were examined for cell cycle kinetics, apoptosis (terminal deoxynucleotidyl transferase assay and p53 protein expression) and immunomorphological features of redifferentiation (MUC1 and DUPAN-2) and endocrine transdifferentiation (insulin, somatostatin, glucagon, neurone-specific enolase) by using immunoperoxidase staining methods. Levels of insulin, transforming growth factor (TGF) beta2, TGFalpha and epidermal growth factor receptor (EGFR) were measured by enzyme-linked immunosorbent assay (ELISA). The vehicle-treated cells served as a control group. RESULTS: When compared with untreated cells, cells treated with 100 nmol/l and 5 micromol/l atRA were observed to show (1) decreased proliferative activity (cpm) as indicated by decreased incorporation of thymidine labelled with hydrogen-3; (2) cell cycle arrest; (3) increased apoptotic activity associated with p53 protein overexpression; (4) upregulated expression of the transdifferentiation and redifferentiation markers; (5) morphological changes indicative of transdifferentiation (increased cell size and appearance of dendrites); (6) decreased production of EGFR; (7) upregulation of TGFalpha and TGFbeta2; and (8) increase in basal and glucose-induced insulin secretion. CONCLUSIONS: Functional endocrine transdifferentiation can be induced in HPAF lines by atRA. Further investigations are mandated to explore the underlying mechanisms of this transdifferentiation and to explore its in vivo extrapolation.

Diabetes mellitus: a risk factor for pancreatic cancer?

The relationship between pancreatic cancer (PC) and diabetes is controversial. While some investigators assume that type II diabetes is a predisposition to PC, recent data argue that diabetes and altered glucose metabolism are a consequence of PC, and yet, the clinical presentation of the altered glucose metabolism in these patients varies considerably. Around 70% of patients with PC have impaired glucose tolerance (IGT) or frank diabetes. Of these, nearly 60% show an improvement of IGT or diabetes after surgery, whereas the rest show only mild or no improvement. It appears that biologically there are three types of PC: (1) PC not associated with IGT or diabetes; (2) PC associated with IGT or diabetes in which the abnormality improves postoperatively; (3) PC associated with IGT or diabetes in which the abnormality does not improve postoperatively. Based on our own studies, we suggest that the reason for impaired glucose metabolism in most patients is the alteration of islet cells either by the carcinogen directly, or by diabetogenic substances released by cancer cells. The extent of the islet alteration (i.e. focal or diffuse) may determine whether the removal of tumor alone can improve the metabolic alteration. The elucidation of the mechanism is of immense importance for providing an early tumor marker and for developing preventative or therapeutic modalities.

MUC4 mucin expression in human pancreatic tumours is affected by organ environment: the possible role of TGFbeta2.

MUC4 is highly expressed in human pancreatic tumours and pancreatic tumour cell lines, but is minimally or not expressed in normal pancreas or chronic pancreatitis. Here, we investigated the aberrant regulation of MUC4 expression in vivo using clonal human pancreatic tumour cells (CD18/HPAF) grown either orthotopically in the pancreas (OT) or ectopically in subcutaneous tissue (SC) in the nude mice. Histological examination of the OT and SC tumours showed moderately differentiated and anaplastic morphology, respectively. The OT tumour cells showed metastases to distant lymph nodes and faster tumour growth (P<0.01) compared to the SC tumours. The MUC4 transcripts in OT tumours were very high compared to the undetectable levels in SC tumours. The SC tumour cells regained their ability to express MUC4 transcripts after in vitro culture. Immunohistochemical analysis using MUC4-specific polyclonal antiserum confirmed the results obtained by Northern blot analysis. Interestingly, the OT tumours showed expression of TGFbeta2 compared to no expression in SC, suggesting a possible link between MUC4 and TGFbeta2. The MUC4 expression, morphology, and metastasis of human pancreatic tumour cells are regulated by a local host microenvironment. TGFbeta2 may serve as an interim regulator of this function.

Immortalization with telomerase of the Nestin-positive cells of the human pancreas.

Cells expressing the neuronal stem cell marker Nestin are present in the human pancreas but the biological role of these cells has yet to be resolved. We report here the establishment with the catalytic subunit of human telomerase (hTERT) of a line of normal human cells representing this cell type. Primary human cells derived from the ducts of the pancreas were transduced with an hTERT cDNA. The infected cells became positive for telomerase, failed to senesce, and were still proliferating after more than 150 doublings. The immortalized cells were positive for the expression of Nestin (at both the mRNA and protein levels) and were found to be free of cancer-associated changes: diploid and expressing wild type p16(INK4a), p53, and K-Ras. An established line of normal human cells representing this cell type should be of great value to help define the biological properties of this novel cell type.

Identification and sequencing of the Syrian Golden hamster (Mesocricetus auratus) p16(INK4a) and p15(INK4b) cDNAs and their homozygous gene deletion in cheek pouch and pancreatic tumor cells.

Previous studies have shown that the p16(INK4a) tumor suppressor gene is inactivated in up to 98% of human pancreatic cancer specimens and 83% of oral squamous cell carcinomas. Inactivation of the related p15(INK4b) gene has also been identified in a number of tumors and cell lines, however, its role as an independent tumor suppressor remains to be elucidated. Chemically-induced tumors in the Syrian Golden hamster (Mesocricetus auratus) have been shown to be excellent representative models for the comparative development and progression of a number of human malignancies. The purpose of this study was to determine the importance of the p16(INK4a) and p15(INK4b) genes in two experimental hamster models for human pancreatic and oral carcinogenesis. First, hamster p16(INK4a) and p15(INK4b) cDNAs were cloned and sequenced. The hamster p16(INK4a) cDNA open reading frame (ORF) shares 78%, 80%, and 81% identity with the human, mouse, and rat p16(INK4a) sequences, respectively. Similarly, the hamster p15(INK4b) cDNA ORF shares 82% and 89% sequence identity with human and mouse p15(INK4b), respectively. Second, a deletion analysis of hamster p16(INK4a) and p15(INK4b) genes was performed for several tumorigenic and non-tumorigenic hamster cell lines and revealed that both p16(INK4a) and p15(INK4b) were homozygously deleted in a cheek pouch carcinoma cell line (HCPC) and two pancreatic adenocarcinoma cell lines (KL5B, H2T), but not in tissue matched, non-tumorigenic cheek pouch (POT2) or pancreatic (KL5N) cell lines. These data strongly suggest that homozygous deletion of the p16(INK4a) and p15(INK4b) genes plays a prominent role in hamster pancreatic and oral tumorigenesis, as has been well established in correlative studies in comparable human tumors. Furthermore, this study supports the comparative importance of the hamster pancreatic and cheek pouch models of carcinogenesis in subsequent mechanistic-, therapeutic-, and preventive-based studies aimed at providing important translational data applicable to pancreatic adenocarcinoma and oral squamous cell carcinoma in humans.

Expression of nerve growth factors in pancreatic neural tissue and pancreatic cancer.

One of the characteristics of pancreatic cancer is its tendency to invade neural tissue. We hypothesized that the affinity of cancer cells for nerve tissue is related to the presence of growth factors in neural tissue and their receptors in cancer cells. Sections of pancreatic cancer and normal pancreatic tissue were examined by immunohistochemistry for the expression of the neurotrophins NGF, BDNF, NT-3, NT-4, and their receptors TrkA, TrkB, and TrkC, as well as the low-affinity receptor, p75NTR. TrkA expression was found in duct, islet, and cancer cells; TrkB was found in the alpha-cells of the islet only. The anti-pan-Trk antibody (TrkB3), which is presumed to recognize all three receptors, immunoreacted with duct and acinar cells in normal tissue and with cancer cells. The staining with TrkC was similar to that of TrkA. The low-affinity receptor p75NTR was expressed in the neural tissue and in scattered duct cells of the normal tissue only. Duct and acinar cells, as well as neural tissue and cancer cells, showed weak to strong immunoreactivity with NGF. NT-3 expression was noted in capillary endothelia and erythrocytes. NT-4 showed specific staining for ductule cells. The expression and distribution of neurotrophins and their receptors suggest their role in the potential of pancreatic cancer cells for neural invasion.

Transdifferentiation of human islet cells in a long-term culture.

It has been established that ductal cells or precursor cells within the ductal tree of the pancreas can differentiate into islet cells. Although islet cells can also form exocrine cells, it is unclear whether they arise from precursor (stem) cells or from mature endocrine cells by transdifferentiation. Using a defined culture medium and technique for islet purification, for the first time we were able to maintain human islets in culture for more than a year. Multilabeling immunohistochemical and immunoelectron microscopic examination of the islets at different days of culture using islet cell markers (antibodies to hormones, neuron-specific enolase, chromogranin A) and ductal cell markers (cytokeratins 7 and 19, carbonic anhydrase II, DU-PAN2, CA 19-9, and MUC1) revealed that endocrine cells gradually transdifferentiate to ductal, acinar, and intermediary cells. Although islet hormone secretion ceased after day 28 in culture, endocrine cells were still detectable at day 60. However, later, all endocrine and exocrine cells were replaced by undifferentiated cells that expressed neuron-specific enolase, chromogranin A, laminin, vimentin, cytokeratin 7 and 19, alpha-1-antitrypsin, transforming growth factor-alpha, and epidermal growth factor receptor. Our data thus show that, under proper conditions, human islets can be maintained in vitro over a long period and that, in the culture condition, islet cells seem to transdifferentiate to exocrine cells and undifferentiated cells, which may be considered pancreatic precursor (stem) cells.

Pacinian corpuscle in the human pancreas.

During our systematic examination of the distribution of cytochrome P450 enzymes in the normal and diseased human pancreas, we observed a Pacinian corpuscle in a serial section of a tissue from a pancreatic cancer patient. We report the histologic and immunohistochemical patterns in this corpuscle and review the literature. The Pacinian corpuscle was situated within the pancreas of a 76-year-old woman with cancer in the head of the pancreas. We could demonstrate immunoreactivity within the corpuscle for the neurofilament protein. neuron-specific enolase, S-100 Protein, and for four cytochrome P450-isozymes. The possible function of Pacinian corpuscles in the mammalian and human pancreas is discussed.

Experimental animal models in pancreatic carcinogenesis: lessons for human pancreatic cancer.

The silent course of pancreatic cancer and its explosive fatal outcome have hindered studies of tumor histogenesis and the identification of early biochemical and genetic alterations that could help to diagnose the disease at a curable stage and develop therapeutic strategies. Experimental animal models provide important tools to assess risk factors, as well as preventive and therapeutic possibilities. Although several pancreatic cancer models presently exist, only models that closely resemble human tumors in morphological, clinical, and biological aspects present useful media for preclinical studies. Because an estimated 70% of human tumors are induced by carcinogens and because a significant association has been found between cigarette smoking and pancreatic cancer, chemically induced models are of particular value. Moreover, in such models the etiology, modifying factors, effects of diets, and naturally occurring products can be studied and early diagnostic, preventive, and therapeutic possibilities sought out. Many of the existing models are described in this review, and the advantages and shortcomings of each model and their clinical implications are discussed.

Increased expression of glutathione S-transferase-pi in the islets of patients with primary chronic pancreatitis but not secondary chronic pancreatitis.

The mechanism of tissue alteration in chronic pancreatitis (CP) is still unclear. Different hypotheses have been discussed, including increasing oxidant stress in the acinar cells, often as a result of exposure to xenobiotics. To evaluate the role of oxidative stress in CP, the authors investigated the expression of the drug-metabolizing phase II enzyme, glutathione S-transferase-pi (GST-pi), in the pancreatic tissue of patients with CP and compared it with the healthy pancreatic tissue from age-matched donors. Pancreatic tissue from patients with secondary CP resulting from ductal obstruction by pancreatic cancer (PC) was also examined. The percentage of cells immunoreacting with anti-GST-pi was counted within 15 randomly selected islets in each slide of the three groups. In all specimens, ductal and ductular cells, and in PC, cancer cells, expressed GST-pi in a moderate intensity. Acinar cells did not stain. Various numbers of islet cells in each of the three groups were stained strongly. More islet cells expressed GST-pi in CP (42%) than in healthy pancreatic tissue (16%, p < 0.001) or PC (17%, p < 0.001). Our results imply an important role of islet cells in the metabolism of substances, which are the substrate for GST-pi, and lend support to the hypothesis of oxidative stress as the cause of CP.

Prevention of pancreatic cancer induction in hamsters by metformin.

BACKGROUND AND AIMS: Our previous study suggested that the known promotional effect of a high fat diet, which in hamsters induces peripheral insulin resistance, is related to a compensatory proliferation of islet cells. The present study was to examine whether the prevention of islet cell proliferation can inhibit the promotional effect of a high-fat diet in pancreatic carcinogenesis. METHODS: Two groups of high fat-fed hamsters were used. One group received Metformin in drinking water for life (HF+Met group), and the other group served as a control (HF group). At the time when the normalization of the plasma insulin level was expected, all hamsters were treated with the pancreatic carcinogen, N-nitrosobis-(2-oxopropyl)amine, and the experiment was terminated 42 weeks later. RESULTS: Although 50% of the hamsters in the high-fat group developed malignant lesions, none was found in the HF+Met group (P < 0.05). Also, significantly more hyperplastic and premalignant lesions, most of which were found within the islets, were detected in the high-fat group (8.6 lesions/hamster) than in the HF+Met group (1.8 lesions/hamster). CONCLUSIONS: The results lend further support on the significant role of islet cells in pancreatic carcinogenesis and may explain the association between pancreatic cancer and obesity, which is usually associated with peripheral insulin resistance.

Combination of tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) and actinomycin D induces apoptosis even in TRAIL-resistant human pancreatic cancer cells.

Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) is a novel member of the tumor necrosis factor superfamily of cytokines that induces cell death by apoptosis. TRAIL has been shown to be effective in almost two-thirds of solid tumors tested thus far, but its effect on pancreatic cancer cells is unknown. We tested the effect of TRAIL on seven human pancreatic cancer cell lines (HPAF, Panc1, Miapaca2, Bxpc3, Panc89, SW979, and Aspc1) in vitro. Of these cell lines, all but Aspc1 showed a significant dose-dependent increase in apoptosis. The apoptotic rate, as detected by a terminal deoxynucleotidyl transferase-mediated nick end labeling assay, was highest in Bxpc3 (71.5%), followed by HPAF (38.0%), Miapaca2 (24.9%), Panc1 (16.1%), Panc89 (15.8%), SW979 (13.9%), and Aspc1 (5.2%). Multiple treatments were more effective than a single treatment and caused a sustained and profound cell death in all but Aspc1 cells. There was no correlation between the effect of TRAIL and the differentiation grade of the cell lines, p53 mutation, or bcl-2 or bax expression. The resistance of Aspc1 cells to TRAIL was not related to the lack of TRAIL receptors. The combination of actinomycin D and TRAIL induced an almost complete lysis of Aspc1 cells, whereas actinomycin D alone had no effect on cell survival but inhibited the expression of the Flice inhibitory protein, which is assumed to play a role in the apoptotic pathway of TRAIL. Thus, the combination of actinomycin D and TRAIL appears to be a promising approach for the therapy of pancreatic cancers resistant to TRAIL.

Early changes in islet hormone secretion in the hamster pancreatic cancer model.

The diabetic state that is seen at a high frequency in association with pancreatic cancer is characterized by elevated plasma levels of several islet hormones and by marked insulin resistance. Both the diabetic state and insulin sensitivity improve after tumor removal by sub-total pancreatectomy. Impaired glucose tolerance has also been found in the hamster pancreatic cancer model, but conflicting data regarding islet function have been reported. In order to further investigate islet function and secretion during early development of pancreatic cancer, we measured the concentrations of insulin, glucagon, somatostatin, and islet amyloid polypeptide (IAPP) in plasma, pancreatic tissue, and secretin-stimulated pancreatic juice at 12 and 27 weeks after the ductal-cell-specific carcinogen, BOP had been used to induce tumors in Syrian golden hamsters. At 12 weeks after BOP, plasma glucagon levels were significantly increased. An exaggerated plasma-glucose response and concomitant hyperinsulinemia were observed at 27 but not 12 weeks after BOP. Plasma IAPP concentrations, but not glucagon or somatostatin, were elevated at 27 weeks. Tissue concentrations of IAPP were substantially reduced in BOP-treated hamsters at 27 weeks. No differences in hormone concentrations were seen in pancreatic juice from the two groups at either of the two time points investigated. The study showed that islet hormone changes accompany the early development of pancreatic tumors in the hamster pancreatic model. The hormone changes and apparent insulin resistance resemble the metabolic changes found in humans with pancreatic cancer.

The patterns of extrainsular endocrine cells in pancreatic cancer.

Abnormal glucose tolerance and frank diabetes mellitus develop in up to 80% of pancreatic cancer patients. Islets within these tumors show a decreased number of beta cells and increased number of alpha cells. The reduced number of beta cells could induce beta cell neogenesis in extrainsular tissue to compensate for the loss of insulin in islets. On the other hand, because the beta cell depletion in pancreatic cancer seems to be the effect of substances released by cancer cells, suppression of extrainsular endocrine cells is expected. We compared the pattern of extrainsular endocrine cells in pancreatic cancer patients with normal pancreas as well as chronic pancreatitis, which is known to be associated with impaired glucose tolerance or frank diabetes. As in the normal tissue, extrainsular endocrine cells were found in chronic pancreatitis and pancreatic cancer. However, in the chronic pancreatitis specimens insulin cells were the predominant cell type, whereas in pancreatic cancer specimens more glucagon than insulin cells were found, although the differences were statistically insignificant. Thus, our results indicate that the alteration of beta cells in pancreatic cancer patients is mainly restricted to the endocrine cells within the islets and that there is no compensatory proliferation of beta cells.

Abnormal differentiation of islet cells in pancreatic cancer.

Pancreatic cancer in many patients is associated with altered glucose metabolism and abnormalities in pancreatic islet hormones at serum and tissue levels. Our previous studies have indicated a tendency of islet cells to differentiate toward ductal cell lineage, but the specificity of these findings for pancreatic cancer was not investigated. In the present study, we examined the immunoreactivity of pancreatic islets to antibodies against tumor-associated antigens DU-PAN-2, TAG-72 and CA19-9 in tissues from the normal pancreas, chronic pancreatitis and pancreatic cancer. Although no immunoreactive islet cells were found in the 12 normal pancreases and 20 chronic pancreatitis patients, 25 of 37 pancreatic cancer tissues showed the expression of these antigens, primarily CA19-9 and TAG-72, where the number of immunoreactive cells varied considerably from case to case. In 4 cases over 50% and in 2 of them more than 75% of the islets showed positive staining of 60-70% of islet cells within each islet. The presence of intrainsular ductular structures expressing the same antigen as the surrounding islet cells suggested transformation of antigen expressing islet cells to ductal cells. All but four islets were within or around the cancer favoring the notion that factors produced by cancer cells are responsible for the altered islet cell differentiation.

Establishment of human pancreatic ductal cells in a long-term culture.

Cultivation and preservation of human pancreatic ductal cells have remained a challenge. With a defined culture medium and refinement of culturing techniques, we have been able to maintain human pancreatic ductal cells without any genetic manipulation in culture for more than 16 months. Freshly isolated ductal fragments were placed on a rocker in M3:5 medium free of collagen for 14 days to remove fibroblasts and endocrine cells before allowing them to attach. The cells produced an excessive amount of mucin and expressed the duct specific cytokeratins (CK) 7 and 19, DU-PAN2, CA19-9, carbonic anhydrase II (CA II), and secretin receptors. During the course of the culture, however, the cells gradually lost the expression of CA II, secretin receptors, DU-PAN2, and CA 19-9 and assumed an undifferentiated phenotype, which showed an upregulation of transforming growth factor alpha (TGFalpha) and epidermal growth factor receptor (EGFR), an increase in the expression of Ki-67, and an increased binding to Phaseolus vulgaris leucoagglutinin (PHA-L) and tomato lectin. These ductal cells present a useful source with which to study physiologic aspects of ductal cells including differentiation.

Inhibition of growth and metastatic progression of pancreatic carcinoma in hamster after somatostatin receptor subtype 2 (sst2) gene expression and administration of cytotoxic somatostatin analog AN-238.

The sst2 somatostatin receptor mediates the antiproliferative effects of somatostatin analogs. The present study demonstrates that stable expression of sst2 in the hamster pancreatic cancer cells PC-1 and PC-1.0 activates an autocrine negative loop leading to an in vitro inhibition of cell proliferation. In vivo studies conducted in Syrian golden hamsters after orthotopic implantation of PC-1.0 cells showed that both tumor growth and metastatic progression of allografts containing 100% of sst2-expressing cells were significantly inhibited for up to 20 days after implantation, as compared with control allografts that did not express sst2. A local antitumor bystander effect was observed after induction of mixed tumors containing a 1:3 ratio of sst2-expressing cells to control cells. Tumor volume and incidence of metastases of mixed tumors were significantly reduced at day 13 post implantation. This effect decreased with time as at day 20, growth of mixed tumors was similar to that of control tumors. After administration of the cytotoxic somatostatin conjugate AN-238 on day 13, antitumor bystander effect observed in mixed tumors was significantly extended to day 20. We also observed that in vitro invasiveness of sst2-expressing PC-1.0 cells was significantly reduced. Tyrosine dephosphorylation of E-cadherin may participate in restoring the E-cadherin function, reducing in turn pancreatic cancer cell motility and invasiveness. This dephosphorylation depends on the tyrosine phosphatase src homology 2-containing tyrosine phosphatase 1 (SHP-1) positively coupled to sst2 receptor. The inhibitory effect of sst2 gene expression on pancreatic cancer growth and invasion combined with chemotherapy with targeted cytotoxic somatostatin analog administration provides a rationale for a therapeutic approach to gene therapy based on in vivo sst2 gene transfer.