Somatostatin, a tetradecapeptide, is a small cyclic peptide. It circulates in the blood in two biologically active forms: somatostatin 14 and somatostatin 28 (429;430), processed from their precursor pre- and pre-pro-somatostatin molecules. Somatostatin inhibits the secretion and action of a number of peptide hormones, neurotransmitters, and exocrine secretions of the GEP axis. It also inhibits a variety of gastrointestinal functions including gastrointestinal motility, gastric acid production, pancreatic enzyme secretion bile secretion and colonic secretions. It inhibits secretion of a number of gastrointestinal and pancreatic hormones such as insulin, glucagon, pancreatic polypeptide, secretin and vasoactive intestinal polypeptide. In addition to it functions as a regulator of neurotransmission and gastric and intestinal hormone secretions, it may control cell proliferation in normal tissues and in tumors(429;431;432). Its clinical use is limited because of its short half-life of 1 to 2 minutes, the need to give it intravenously and the multifarious effects and the rebound of secretions when it was stopped (433). Development (433) of its potent, long-acting octapeptide analogue (Sandostatin, octreotide acetate) with a half-life of over 100 minutes was a breakthrough for clinical application. Other analogues such as somatuline (BIM 23014 C) are being investigated. Thus far, their effectiveness and toxicity rates appear to be similar to those of octreotide. Several aspects of treatment of GEP neoplasms, including symptom reduction, hormone suppression, tumor growth, and survival, are discussed here.

Somatostatin-14 and somatostatin 28 act through high affinity G protein coupled membrane receptors. Five somatostatin receptor (SST) subtype genes have been cloned and characterized, SST1-5 (434) and are localized on different chromosomes (435). Two forms of the SST2 (a and b) are generated through alternative splicing (436), (435). A number of systems are activated after binding to the respective receptors: 1. inhibition of adenylate cyclase; 2. inhibition of calcium channels; 3. stimulation of phosphotyrosine phosphatase; 4. inhibition of MAP kinase activity(437), (435;438). The effects on adenylate cyclase and calcium channels are important for hormone and intestinal secretion and those on the phosphotyrosine and MAP kinases for cell growth and proliferation (431;432).

Somatostatin target tissues express multiple SST subtypes although individual tissues may express different subtype uniquely. Pancreatic islet cells express all five SSTs (439;440). However in these cells the SST 1, 2 and 5 subtypes are the most abundant with pancreatic b cells expressing predominantly SST1 and 5, and a cells expressing SST2, δ cells SST 5(440).

Tumors arising from somatostatin target tissues often express a high density of SSTs.(441-445). These tumors include pancreatic endocrine tumors, carcinoids, paragangliomas, pheochromocytomas, medullary carcinomas of the thyroid, in addition to pituitary prolactinomas and non endocrine tumors such as the breast and malignant lymphomas (441;446). This can be determined immunohistochemically or by in situ hybridization and RNAase protection assays and RT-PCR (143;447). The majority of these tumors express multiple subtypes of SSTs but there is considerable variation between subtypes expressed by different tumors of the same class, and even different metastases deriving from the same tumor (439;448-450). Endocrine pancreatic and gut tumors have SST2 more than 80% of the time (443;451-453).

The five subtype of SSTs bind somatostatin 14 and somatostatin 28 with high affinity. The SST4 however does not bind octapeptide analogues octreotide and lanreotide whereas the SST2, 3 and 5 display high, moderate and low affinity for these analogues. The predominant expression of SST by endocrine tumors of the pancreas and GI tract forms the basis of use of these analogues for the treatment of the expression of the endocrine tumor (454),(204) . The high density of the SST subtype in these tumors predicates the use of radiolabel analogs for the localization of tumors and their metastases and indeed is predictive of response to therapy, and parallels the presence of receptors identified immunochemically (147;455-459). This knowledge is important in deciding on the appropriate from of therapy.

Figure 1

An interesting recent discovery is that the SST receptors may form heterodimers with other G protein coupled receptors such as the dopamine D2 and the u opiod receptor (MOR-12) resulting in a novel receptor state with properties different from the parent receptors (460;461). This predicts that future analogs may become hybrid molecules with the ability to bind and activate these hybrid receptors and expand the limited capability in certain tumors or even overcome the apparent tachyphyllaxis that seems to occur in many of the tumor syndromes.

Figure 2

As discussed above the short half life, the tendency to rebound , the need to give somatostatin intravenously by infusion, created the need for synthesis of analogs with longer half lives which could by administered by routes other than IV and that suppressed secretions for protracted periods without a tendency to rebound. The first such analog was octreotide (Sandostatin, Novartis, Basel, Switzerland). Its elimination half life after SC administration is of the order of 2h and rebound hormone secretion dose not occur (462) Octreotide does not bind to all the SSTs but preferentially binds to SST2 and 5 with high affinity (438). Other cyclic analogs with almost similar affinity and specificity include lanreotide (Somatuline, Ipsen, Biotech, Paris, France) have subsequently been developed (433). The slow release deport intramuscular formulation of octreotide ( Sandostatin LAR, Novartis Pharma, Basel, Switzerland) has to be administered once every 4 weeks and lanreotide ( Somatuline PR, Ipsen Biotech, Paris France) once every two weeks. A new slow release of Lanreotide, Lanreotide Autogel (Ipsen Biotech, Paris, France) has been introduced in Europe and can be given by deep subcutaneous injection every 4 weeks.

In most patients with carcinoid tumnors and pancreatic endocrine tumors treatment with octreotide induces a rapid clinical response in the symptotology such as diarrhea, dehydration, flushing attacks, hypokalemia, peptic ulceration and the necrotic skin lesions . Unfortunately many of these patients show desensitization of the inhibition of hormone secretion by octreotide within weeks to months. In a series of 57 patients with carcinoid, octreotide therapy was ended in 23 patients after periods ranging from 1 week to 12.5 months, median 4 months, whereas the remainder could be controlled for periods up to two and one half years. The estimated mean duration of response of the whole group was 1 year (463).These were early studies and no attempts were made to adjust dosage, use periods free from the analogs for reduction of tachyphyllaxis or give the agents by continuous infusion as has been shown to be successful in some recalcitrant cases and certainly is the best form of therapy for acromegaly (464). Potential mechanisms of the tachyphyllaxis and resistance to the action of somatostatin are :. receptor down regulation, a decrease in the number and affinity of the SSTs, desensitization due to uncoupling of the receptor from the G protein second messenger, differences in the degree and type of receptor expression in the tumor metastases, or even growth and development of clones resistant to the action of somatostatin, resistance of the target hormone secretory mechanisms to somatostatin or their growth promoting properties for these tumors e.g. growth hormine or gastrin, and finally mutations in the SST genes and their functional peptide receptors (465).

Octreotide has a potent action in reducing symptoms in certain neuroendocrine tumors. Detection of somatostatin receptors by octreotide scintigraphy correlates well with the predicted response to treatment with octreotide (155). In carcinoid tumor, flushing is reduced in most patients (205;206;466). The acute effects on water and electrolyte transport are a reversal from a secretory to an absorptive state, thus normalizing transport across the proximal intestine (204). Long-term responses of diarrhea, however, differ in different reports: 9 of 14 patients with endocrine diarrhea responded to treatment in one report,(211) in contrast to 19 of 25 in another (205). Diarrhea in VIPoma improves 95% of the time (466). This difference might result from the involvement of different peptides causing diarrhea and different mechanisms. In five of our patients with gastrinoma, the presenting symptom was diarrhea, which improved, as did 65% of 26 reported cases(466). Diarrhea in 16 patients with glucagonomas also improved uniformly (466;467). Diarrhea also improves in all patients with gastrinoma syndrome, and abdominal pain can be relieved in most patients (98). There is, however, the possibility of a rebound in symptoms and/or hormonal values during therapy. The mechanism of this is not clear, but it might involve accelerated enzymatic breakdown of octreotide and/or ligand-induced changes of somatostatin receptors on the target cell, preventing internalization of the hormone receptor complexes or a gradual adaptation of the target cell to the octreotide effect, as proposed by Koelz and colleagues (468).

Wheezing, as one of the symptoms in carcinoid syndrome, can be reversed by octroetide, and spirometric improvement in lung function has been documented (211). In another patient who had severe proximal myopathic muscle weakness, clinical and electromyographic improvement occurred with octreotide treatment (211). The arthropathy of carcinoid, which may be SP mediated, also improves (389). Hypoglycemia with insulinoma responds erratically because of unpredictable effects on food absorption, suppression of glucagon, and insulin (204). Of 15 patients, 50% improved, and 30% got worse (466). The necrolytic migratory erythema of glucagonoma clears in only 50% of the cases. (469)

Octreotide inhibits hormone secretion in some malignant GEP tumors. The most sensitive of these seems to be VIPoma, where lowering of VIP circulating levels parallels relief of symptoms (204;207). Gastrin levels, however, are not equivalently lowered with octreotide. One study of eight patients with gastrinoma showed that octreotide decreased gastrin levels in five patients by a mean of 76% of baseline (98). In worldwide pooled data (n = 26), 70% of patients are reported to respond. ACTH overproduction heralds unresponsiveness to the drug. Glucagon levels seldom decrease (466). Although the overall 5-HIAA level is significantly lower after octreotide treatment, (205) blood serotonin level does not differ significantly (206;470). There is an overall reduction of 5-HIAA in 58% of patients (466).

Perioperative Management

Carcinoid and other GEP tumors can be a major therapeutic problem perioperatively, when vast quantities of active peptides are released into the circulation from manipulation of the tumors. Octreotide is an effective suppressor of release and action of peptide hormones during surgery (278). Profound refractory hypotension in carcinoid syndrome can be rapidly reversed by octreotide, (203) as can gastric acid secretion and fistula drainage (278).

Tumor Growth

While expression of SSTs is essential for the control of hormone and intestinal secretions by somatostatin and its analogs, somatostatin has been shown to inhibit normal and tumor cell growth. Induction of G1 arrest and apoptosis has been demonstrated in a number of tumor cell models and involves a number of the SST receptor subtypes (471-473) , (474), (475-481). Reports of tumor regression or even infarction support the notion that somatostatin has antitumor activity aside from controlling secretions (482). Based upon clinical observations that administration of somatostatin during angiography of metastases to the liver, Cho and Vinik postulated that it had vasoconstrictive or anti-antigenic properties (483) and indeed it is now known that these drugs are anti-angiogenic in malignant vasculature, inhibiting vascular growth factors and are anti apoptotic for malignant cells (484-492).

Figure 3

Tumors that have SST 2 or SST 5 receptors can be visualized in vivo after the injection of radiolabeled octapeptide analogs of somatostatin. The technique was first developed using the radiolabeled somatostatin analog 123I-Tyr3 octreotide (493). Unfortunately this analog was difficult to prepare, cost was prohibitive, it accumulated in the liver gallbladder and ducts and gastrointestinal tract forcing the development of newer and better analogs. The most widely used for SST scintigraphy is now ¹¹¹In-pentreotide (¹¹¹In-DTPA ⁰ octreotide, Octreoscan, Tyco Healthcare, Mallinckrodt, St Louis USA) (146). Apart from ¹¹¹In-pentreotide, ¹¹¹In-DOTA⁰ lanreotide can also be used (494;495). Of greatest importance in this regard is the fact that a negative scan predicts absence of the appropriate SSTs and does not augur well for a tumor response. In clinical practice a negative scan suggests that alternative therapies should be sought. It is the relationship between tumor mass identifiable on CT, MRI or even PRET and the concordance with the somatostatin scan that suggests that the tumor is indeed receptor positive and likely to respond. Dissociations between the Octreoscan and other modalities of imaging suggest that the tumor is devoid of the needed SSTs and will not respond. Thus clinical decision making is simplified by this approach (496).

Octreoscans may not be universally equivalent. Granberg et al(497) from the University Hospital in Uppsala identified 28 patients with histologically verified bronchial carcinoids and prospectively carried out Octreoscans and Ct Scans of the chest . Altogether 71% had octreoscan positive tumors. The primary tumor was positive in 13/16 whereas the CT was positive in 15/16. CT failed to localize the tumor in 1 octreoscan positive patient. Intrathoracic metastases were visualized by Octreoscan in 7/9 patient and by CT in 8/9 patients. Of the liver metastases octreoscan was positive in 9/14 (64%) and of bone metastases it was positive in 9/10. The authors conclude that octreoscans are useful for following patients with bronchial carcinoids but that CTs are better for identifying the primary tumor as well as metastases to the liver.

Because GEP tumors grow slowly, it is hard to assess the effect of treatment on tumor growth. Long-term CT monitoring, however, has shown shrinkage of liver metastasis in certain patients with carcinoid and other GEP tumors (205;216). In 85 carcinoid tumors, no change was found with doses under 50 mg/d, but Kvols and colleagues¹¹² reported a decrease in size in 17% of patients using higher doses (1,500 mg/d). VIPoma, glucagonoma, and gastrinoma generally do not change in size (466), although a tumor infarction has occurred in VIPoma (278). Tumor metastases to bone may occur despite apparent control of the primary tumor or liver metastases (204;498).

It might be feasible to use the peptide in both peptide therapy as well as in SST targeted radiotherapy. Once bound to its SST the agonist SST complex is internalized as are many G protein coupled complexes (465;499-503). The SST subtypes have a differential rate of internalization which may account for differences in the susceptibility of different tumors ion different target tissues. The SST1 show low agonist-induce internalization whereas the SST₂, SST₃, SST₄, and SST₅ are more efficient in this respect (438;465). When ¹¹¹In-DTPA^o octreotide is administered to patients with carcinoid tumors it is taken up, bound to the plasma membrane, internalized and appears in the cytoplasm and in the cell adjacent to the nucleus. Proximity to the nucleus is important for this short range Auger electron emitting radioisotope to exert its cytotoxic effect(504). Thus the predominant expression of SST₂ receptors on these tumors and the avidity of this particular receptor for octreotide make it particularly suitable for application of SST targeted therapy. However, the Auger emitter ¹¹¹In has a low tissue penetration and stable a or b emitting isotopes could not be developed. As a result novel compounds were developed allowing the use of more powerful isotopes e.g. [DOTA⁰, Tyr 3] octreotide allowing stable binding with the b emitter yttrium 90 (Octreother, Novartis, Basel Switzerland). In addition ¹¹¹In-DOTA0] lanreotide and ⁹⁰Y-DOTA⁰] lanreotide can be used for advanced tumors containing the SST₂ and SST₅ receptors (450;505-509). While several mechanisms determine the efficacy of these radiotherapeutic drugs which are not unlike the biological effects of somatostatin analogs, there is the added need for stability of the radioligand, the density of SSTs on the tumor, the affinity for the ligand, the efficiency of internalization and the trapping of radioisotopes within the tumor cells, tumor and the mass of injected peptide(465;510;511).

Because of this unique aspect of these tumors and their tissue receptor expression and the selectivity of currently available analogs, it was expected that new analogs with different specificities and would be found that bound to SSTs other than SST2 and SST 5. A new- so called universal somatostatin analogue has been found and named SOM 230, which has high affinity for SST _{1,2,3 and 5}. It is currently being evaluated in phase 1-111 trials (512;513) and shows promise for acromegaly and the reduction of IGF-1 levels. New drugs with multi receptor binding capabilities and analogs that form homo or hetero dimmers may yet offer greater potential of uptake and internalization by these tumors. Possibly this will yield improved results in terms of desensitization (460;514;515). .Hybrid molecules, for example somatostatin –dopamine, have a high affinity for both somatostatin and dopamine D₂ receptors and may be particularly useful for prolactin secreting tumors. While this peptide radiotherapy seems exciting it has yet to fulfill its promise. In the USA today it is only available at select centers doing trials and has not yet been incorporated into any schems whereby patients with these tumors are being treated. Other alternative approaches such as gene therapy with targeted ablation of an SST subtype over-expressed on a tumor or induction of expression thereby increasing the likelihood of response to an analog are in the future.

It has recently become clear that endocrine cells are not the only tissues that express the SST receptors. There is now evidence that there is differential expression of SSTs on low and high grade astrocytomas (516) and although this did not show predictive value for survival time it does raise the possibility of developing a new therapy for this notoriously difficult tumor. Even in pancreatic cancer there appears to be a difference in the expression of SST2 in the cancer and its adjacent tissue, which might be reason for the differences in therapeutic effects of somatostatin on these cancers. Interestingly these tumors express DCP4, p53 and ras genes reciprocally with the SST2 receptor and may thus participate in pancreatic angiogenesis. (517) Analogs that target the SST2 receptor may compromise angiogenesis so essential for the growth and proliferation of these tumors. DOTA Tyr3-octreotide (DPTATOC) and DOTA 0 Tyr 3-octreotate, bound to beta emitting radioisotopes, are suitable for therapeutic use but this is still in the research arena. From the point of view of radiometabolic therapy directed towards the destruction of micro metastases, pharmacological research has provided a host of potential candidates : octreotide LAR, lanreotide, vapreotide, BIM-23244, BN 81644, PTR 3173, BIM 23A387, SOM-230 with different pharmacokinetic properties and receptor binding characteristics (518). The most developed of these is perhaps SOM 230. It is a cyclic hexapeptide which is metabolically stable and has a terminal half life in animals of 23h. It is about 40 times as potent as somatostatin in the inhibition of growth hormone and by continuous infusion causes 75% reduction in IGF-12 levels compared with 28% by somatostatin itself. It binds to the SST 1,2,3 and 5 receptor subtypes, but has less potency on insulin and glucagon secretion and does not alter glucose tolerance. It seems that it will be a useful analogue for acromegaly but the question of its added advantage in carcinoid and other tumors of the gastroenteropancreatic axis remain to be resolved .(512)

Both somatostatin and its analogs as well as interferon alpha have been shown to control hormonally active functional neuroendocrine gastroenteropancreatic tumors. Faiss et al (519) carried out a prospective randomized multicenter trial on the anti-proliferative effects of lanreotide, interferon alpha and their combination for metastatic neuroendocrine tumors in 80 therapy naïve patients with histologically proven neuroendocrine tumors (primary location: foregut, n=36, midgut, n=30, hindgut, n=3, unknown, n=11, functional , n=29, non-functional, n=51). Patients were randomly treated with lanreotide( 1mg tid, SC) or interferon alpha ( 5X10-6 units tiw, SC) or the combination of the two. . All patients had disease progression prior to entry into the study verified by imaging studies. Twenty five patients were treated with lanreotide, 27 received interferon α and 28 patients received the combination. Partial tumor remission was seen in 4 patients , 1 on lanreotide, 1 interferon α and 2 patients who received the combination. During 12 months of therapy, stable disease was observed in 19 patients (7 on lanreotide, 7 on interferon α and 5 on the combination) whereas tumor progression occurred in 14/25 patients on lanreotide, 15/27 on interferon α and 14 on the combination. Side effects leading to interruption were more frequent in the combination group than either of the individual groups. The authors of the “International Lanreotide and Interferon Alfa Study Group (519) conclude that lanreotide and interferon α have comparable anti-proliferative effects in the treatment of metastatic neuroendocrine tumors of the Gastroenteropancreatic axis...

Figure 4

Figure 5

Response rates were however lower than those previously reported in non randomized studies and no attempts were made to escalate dosage or to deliver the somatostatin by continuous infusion. It would seem that the combination may yield slightly greater advantage. No attempt was made in the study to determine if those failing on lanreotide could be rescued with interferon. Similar results were found for the functional and non functional tumors but however there are very marked differences in clonal responsiveness and the presence or expression of SSTs, so the study fails to address the true response rate in receptor positive people and if dosage had been maximized. It nonetheless suggests that in the presence of failed optimized therapy with a somatostatin analogue there is room for the introduction of interferon and that the side effects are not. Figure 18 indicates just such an example. Patient MT had a very good response initially to LAR, 30 mg /day with a fall in serotonin values to near normal. However, over the next 9 months serotonin values began to rise and Sandostatin 1500ug/day was added to the LAR without much success. However the addition of Interferon alpha 4M units 5X/week caused a dramatic reduction in serotonin values. Cases such as these indicate that the use of Inteferon may represent a fall back position when one is losing ground with LAR and somatostatin.

The role of octreotide in the treatment of GEP tumors is still well established. Because of the clear evidence of symptomatic relief (e.g., flushing, wheezing, diarrhea), it has established a place in treatment of such tumors both pre- and postoperatively. Perioperative use can prevent fatal episodes of rapid, extreme increases of hormones in the circulation. There is enough evidence for the control of tumor growth that primary treatment of select metastatic tumors, with proper monitoring of tumor growth, is recommended.