New Perspectives in the Treatment of Neuroendocrine Tumours

Somatostatin Analogues

The biological activity of natural somatostatin is based on the inhibition of hormone secretion, in the suppression of the release of insulin-like growth factors, in the negative control of angiogenesis, and in the enhancement of apoptotic processes in the tumour tissue. Physiologically, somatostatin is a potent secretion inhibitor of several hormones derived from the hypophysis, pancreas and gastrointestinal tract (growth hormone, insulin, gastrin, cholecystokinin, etc.); in addition, it regulates different cell activities such as growth, gastro-intestinal motility and fluid secretion (9).

All these activities are mediated by the interaction of somatostatin with a series of five receptors (SSTRs), encoded by five different genes belonging to the class of receptors linked to transmembrane G-proteins, able to inhibit cAMP (9).

Therapeutic activity is achieved through the interaction with two out of five SSTRs and, more precisely, with subtypes 2 and 5 for which there is the highest affinity (9).

Octreotide and lanreotide are the two somatostatin analogues exploitable by injection. They proved highly effective in the symptomatic treatment of NETs, being able to control the symptoms related to carcinoid syndrome in up to 90% of cases. Reduction of tumour biochemical markers is reached in up to 70% of cases (8, 10-13).

A recent randomized phase III study (PROMID study) (14) has demonstrated that in patients with advanced midgut NETs, octreotide LAR compared to placebo induces an advantage in terms of progression-free survival (14.3 months vs. 6.0 months, p=0.000072).

An analogue named SOM 230 (pasireotide), able to bind four out of five receptors (all except SSTR4) is presently under clinical development. On account of this characteristic, SOM 230 is potentially more effective as compared to the analogues presently available, since it may also be active in tumours unresponsive to somatostatin analogues, which does not bind subtypes 2 and 5. Clinical trials are ongoing to assess the true efficacy of SOM 230 in different types of NETs (9).

On the whole, all somatostatin analogues have a good safety profile even in cases of long-term treatment. Adverse events occurring in 25-50% of cases usually are mild to moderate in severity and do not require treatment discontinuation, being most frequently represented by development of gallbladder stones, pain at the site of injection, abdominal colic, flatulence, nausea, asthenia and reduced tolerance to glucose.

Chemotherapy

Several chemotherapy agents have been employed either as single-agent or in combination in the treatment of advanced stage NETs, such as streptozotocin, doxorubicin, 5-fluorouracil, cisplatin, etoposide and dacarbazine. Recently, some new chemotherapeutic agents have become available, such as temozolomide, oxaliplatin, capecitabine, irinotecan and gemcitabine. Taking into account the lack of clinical studies in this setting, there are currently no indications for chemotherapy in adjuvant or neoadjuvant treatment. The available clinical results strongly vary on the basis of the utilized agents and on the prognostic characteristics of NETs. In particular, the variable most likely predictive of responsiveness is a very high proliferative index (>15%) (5).

Well-differentiated GEP-NETs. Single-agent chemotherapy with streptozotocin yielded a tumor response rate of 36-42%, but these early studies can be criticized with respect to the rough methods of interpreting morphological responses. Other monotherapies, including chlorotozotocin, doxorubicin, 5-FU and dacarbazine have been used, but were criticized due either to the high toxicity rate or lack of objective response. Monotherapy strategies have been universally replaced by combination chemotherapy protocols. As can be seen in Table I (4), many combinations have been used, with streptozotocin, 5-FU and anthracyclines making-up the cornerstone of the tested regimens. In well-differentiated pancreatic endocrine tumors the results obtained by Moertel et al. using streptozotocin and doxorubicin, so far, have not yet been improved, with a 69% objective response rate and a median survival of 26 months; these results should be compared with an objective response rate of 45% for 5-FU and streptozotocin. The same group had previously obtained better results with 5-FU in combination with streptozotocin in a phase III trial, in comparison with streptozotocin monotherapy (6-7). While no group has managed to achieve the same response rates, objective responses of 36-55% have been established using streptozotocin plus doxorubicin, with the exception of one study where a response rate of 6% was reported in a group of 16 patients. The authors questioned the reliability of Moertel et al.'s earlier studies, especially their methods of measuring responses. However, three recent studies have reported good response rates using well-defined criteria for recruitment and evaluation. Strosberg et al. reported that the combination of capecitabine and temozolomide is associated with a high and durable response rate in metastatic endocrine carcinomas of the pancreas, superior to those observed with streptozotocin-based regimens (26). As for well-differentiated GEP tumors of the pancreas, single-agent regimens have been largely disappointing in GEP tumors of midgut origin, with objective response rates of <25% and response durations rarely exceeding three months. In 1979 Moertel et al. combined 5-FU with streptozotocin for midgut carcinoids, yielding a response rate of 33%. Later studies using the same combination have failed to reproduce these results (Table I). Therefore, other drug combinations have also been examined, but apart from a 40% objective response rate for patients with midgut carcinoids treated with doxorubicin and streptozotocin in a phase II study, no other reliable cytotoxic regimen has been found for patients with advanced or metastatic disease of midgut origin (15-27).

Poorly-differentiated GEP-NETs. Standard treatment of patients with advanced poorly-differentiated GEP tumors has largely been based on protocols containing etoposide and cisplatin (Table II); such patients are rarely sensitive to combination therapies with streptozotocin, dacarbazine and 5-FU. Other studies were conducted with cyclophosphamide, dacarbazine and vincristine, recording an unsatisfactory response duration. A study involving the combination of 5-FU, epirubicin and dacarbazine, had a lower response rate compared to the regimen of cisplatin and etoposide. While tumor response rates are often good (42-65%), the duration of response rarely exceeds 10 months and median survival is in the order of 15 months. The guidelines state that a cisplatin plus etoposide regimen is indicated; however XELOX (capecitabine and oxaliplatin) or FOLFOX (5-FU, lederfolin and oxaliplatin) chemotherapy regimens (15) or even the combination of CDDP with a molecular-targeted therapy, can be considered as backup. New options are required for the treatment of these patients (28-36).

Thoracic NETs. NETs of the thorax include both bronchial and thymic NETs. No adjuvant chemotherapy or chemoradiation is recommended for well-differentiated tumors. G1/G2 bronchial NETs are generally less responsive to chemotherapy than SCLC. However, platinum-based regimen may be considered and have reported activity in patients with more aggressive/intermediate-grade tumors. The use of various chemotherapeutic agents (doxorubicin, 5-FU, dacarbazine, cisplatin, etoposide, streptozotocin and carboplatin) in the treatment of lung carcinoids has yielded minimal (20-30%), mostly short-lasting results, and an effective chemotherapeutic regimen for unresectable disease is still lacking. Combination chemotherapies for carcinoids are usually platinum and streptozocin-based. Due to the low response rates for chemotherapy in BP-carcinoids combined with serious side effects, the indication to use currently available chemotherapeutic regimens is limited (25). Results from a published phase II trial suggest antitumor activity with single-agent temozolamide for well-differentiated NETs (26). In small studies, large-cell NETs have been shown to have a low and partial response rate to preoperative or postoperative chemotherapy but therapy prolongs survival in lower-stage disease (27, 28). The standard-of-care for limited-stage SCLC includes early thoracic radiotherapy combined with cisplatin and etoposide. Extensive-stage disease is usually treated with chemotherapy with etoposide and a platinum compound, considered as the reference treatment for inoperable poorly-differentiated NETs (23).

The New Drugs

Monoclonal antibodies – bevacizumab. Bevacizumab (Avastin®) is a humanized murine monoclonal antibody directed against Vascular Endothelial Growth Factor (VEGF), utilized in the treatment of several neoplasms with antiangiogenic purposes. A randomized phase II study has demonstrated that in patients with advanced carcinoids bevacizumab (15 mg/kg q 3 w) compared to interferon-2β has an advantage in terms of progression-free survival. Another randomized phase II trial carried out in patients with advanced carcinoids on treatment with octreotide and with progressive disease showed a benefit of bevacizumab over IFN-α in terms of objective responses and progression-free survival (37-41).

Tyrosine kinase inhibitors (TKIs). Sunitinib malate (SU-11248, Sutent®) is a selective inhibitor of some tyrosine kinases (TKI), such as Vascular Endothelial Growth Factor Receptor (VRGFR)1, 2 and 3, Platelet-Derived Growth Factor Receptor (PDGFR), c-KIT and REarranged during Transfection (RET) gene. It exerts a dual antitumour activity either as angiogenesis inhibitor or as antiproliferative agent through a direct effect on tumour cells. The clinical activity of sunitinib has been evaluated in phase II and III studies in patients with pancreatic NETs. In a randomized, double-blinded, multinational, phase III trial, continuous treatment with oral sunitinib at 37.5 mg/day significantly prolonged the median progression-free survival time (primary end-point) by approximately two-fold relative to placebo, in adults with locally-advanced and/or metastatic, well-differentiated pancreatic NETs. Sunitinib was also associated with a significantly greater objective tumour response rate than placebo, although limited data from an updated analysis demonstrated no significant difference between the treatment groups, in terms of median overall survival. On account of the drug-related side effects, of the difficulty to identify the optimal duration of treatment and of the slow-growing feature of NETs, further investigations are needed to clarify the possible role of sunitinib in this disease (42, 43).

m-TOR pathway inhibitors. Clinical investigations aiming at evaluating the efficacy of everolimus (RAD001), both as single-agent and in combination with octreotide LAR, in patients with advanced NETs have achieved interesting results. An international, multicenter, phase II study (RADIANT1), carried out in a significantly representative population of 115 patients with pancreatic NETs has confirmed the efficacy of RAD001, employed as single-agent and in combination with octreotide, in terms of disease control. The preliminary results of the RADIANT2 study in patients with GEP and lung NETs with carcinoid syndrome have not reached the primary end-point (PFS). Instead, the RADIANT3 study confirms the efficacy of RAD001, compared to placebo, in pancreatic NETs (44-46). In conclusion, following recent efficacy data, the therapeutic algorithm, particularly in pancreatic NETs should include the use of biological drugs, such as sunitinib and RAD001, after failure of somatostatin analogues.

There are currently several studies underway to assess efficacy and safety of some other new molecules.

Cabozantinib is a potent dual inhibitor of the MNNG HOS Transforming gene (MET) and VEGF pathways, designed to block MET-driven tumor escape. It is under investigation in pancreatic neuroendocrine tumor in a phase II study. (Trial n°: NCT01466036)

Panzem (2-methoxyestradiol, 2ME2) nanocrystal dispersion (NCD) is derived from estrogen and works by suppressing tumour growth and blocking the formation of new blood vessels that feed tumours. It is administered orally with recombinant human monoclonal antibody against VEGF (bevacizumab) which is administered intravenously, to patients with locally-advanced or metastatic carcinoid (Trial n°: NCT00328497).

Axitinib is a small-molecule TKI. It inhibits multiple targets, including VEGFR1, 2, 3, PDGFR, and cKIT (CD117). A phase II study will assess efficacy and safety of this drug in carcinoids (Trial n°: NCT01435122).

EPO 906 is a potent member of a new class of microtubule-stabilizing cytotoxic agents, known as epothilones. A phase II study will examine if EPO906, given by intravenous infusion, is effective in shrinking tumors and preventing the growth of cells (Trial n°: NCT00050349).

Pansierotide LAR is a somatostatin analogue that binds for each of the five known SSTRs. A randomized, multicenter, phase III study is comparing the efficacy of pasierotide LAR and octreotide LAR in patients whose disease-related symptoms are inadequately controlled by currently available somatostatin analogues (Trial n°: NCT00690430).

Vatalanib is an oral tyrosine kinase inhibitor targeting VEGFR1, 2, and 3. In a phase II trial vatalanib together with octreotide are being studied to understand if they work in treating patients with progressive neuroendocrine tumors (47). (Trial n°: NCT00627198)

Radiotherapy of NETs

The treatment of NETs with radiopharmaceuticals is possible today, both for radiopharmaceuticals targeting receptors overexpressed by these tumours and also with radiopharmaceuticals acting on the cell metabolism. Different receptors have been investigated as a target of the radioisotope, however, so far the SSTRs seem to be the best option. The radiopharmaceuticals targeting SSTRs are based on three components: a peptide, a chelator and a radionuclide. The first somatostatin analogue synthesized was octreotide; the substitution of phenylalaline at position 3 with a tyrosine-residue produced Tyr3-octreotide (TOC). This increased the affinity for SSTR2. Replacing the C-terminal threoninol with threonine resulted in the synthesis of Tyr3-octreotate (TATE), which has been shown to have much higher affinity for SSTR2 when compared to octreotide. Many other analogues have been developed, by substituting chemical groups with the aim of enhancing the affinity of the analogues for the receptors. Among them, one of the most recent is Nal3-octreotide (NOC), obtained by substituting a naphtyl-alanine in position 3. An important part of these radiopharmaceuticals is the radioisotope which is bound to pepdide through the chelator diethylentriaminepenta-acetic-acid (DTPA) or tetra-azacyclododecanetetra-acetic-acid (DOTA).

The most relevant radiopharmaceuticals that should be cited for peptide receptor radionuclide therapy (PRRT) are (Table III): ¹¹¹In-pentetreotide (Octrescan®), ⁹⁰Y, DOTA (Tyr3) TOC, ¹⁷⁷Lu-DOTA (Tyr3) TATE.

Valkema et al. treated 26 patients with GEP NETs tumours with high doses of ¹¹¹In-DTPA octreotide, receiving a total cumulative dose of more than 20 GBq. The results were: 8% partial response (PR), 58% stable disease (SD). In other studies patients were treated with high cumulative activities (up to 36.6 GBq) and 17% of them had PR, with 58% SD. In all studies, the most common toxicity was bone marrow suppression (48).

Therapy with ⁹⁰Y-DOTATOC was performed by Otte et al. who treated 29 patients with GEP-NETs using a dose-escalating scheme of four or more cycles of ⁹⁰Y-DOTATOC, up to a cumulative dose of 6.120±1.347 MBq/m². The results were: 24 patients had SD, two had PR and three had progressive disease (PD). (49). Bodei et al. published data of a phase I study in 21 patients with GEP NETs. Cumulative total doses given in two cycles ranged from 5.9-11.1 GBq. The results were: 29% PR, with a median duration of response of 9 months. The same group evaluated the objective response of 141 patients with various types of NETs, treated with doses higher than 7.4 GBq of ⁹⁰Y-DOTATOC (cumulative activity: 7.4-26.4 GBq) divided into 2-16 cycles. An overall clinical benefit (CR+PR+SD) was observed in 76% of patients (50).

The most important study of the use of ⁹⁰Y-DOTATOC is a phase II study by Imhof et al., who investigated the efficacy, survival and toxicity of such therapies in 1,109 patients. Out of these patients, 378 (34.1%) experienced morphological responses; 172 (15.5%), biochemical responses; and 329 (29.7%), clinical responses. During a median follow-up of 23 months, 491 patients (44.3%) died. Longer survival was correlated with each: morphological (hazard ratio (HR)=0.46; 95% Confidence Interval (CI)=0.38-0.56; median survival, 44.7 vs. 18.3 months; p<0.001), biochemical (HR= 0.75; 95% CI=0.59-0.96; 35.3 vs. 25.7 months; p=0.023), and clinical response (HR=0.68; 95% CI=0.56-0.82; 36.8 vs. 23.5 months; p<0.001). Overall, 142 patients (12.8%) developed grade 3 to 4 transient haematological toxicities, and 103 patients (9.2%) experienced grade 4 to 5 permanent renal toxicity. Multivariable regression revealed that tumoral uptake in the initial imaging study was predictive of overall survival (HR=0.45; 95% CI=0.29-0.69; p<0.001), whereas the initial kidney uptake was predictive for severe renal toxicity (HR=1.59; 95% CI=1.17-2.17; p=0.003). (51) The advantages of using ¹⁷⁷Lu DOTA-TATE have the best tumour/kidney, spleen and liver uptake ratio which allows higher tumour-absorbed doses without major effects on the dose-limiting organs, the longer residency time of ¹⁷⁷Lu DOTA-TATE in tumours and the gamma emission of ¹⁷⁷Lu (52). Most of the studies using ¹⁷⁷Lu-DOTA-TATE have been performed by Kwekkeboom et al. who propose ¹⁷⁷Lu octreotate as the radiolabeled somatostatin of choice when performing PRRT. In 2003 they assessed the effects of ¹⁷⁷Lu-DOTA-TATE in 34 patients with GEP tumours. The results were: 3% CR, 35% PR, 41% SD and 21% PD. Following this study they treated 131 patients with GEP tumours with a cumulative dose of 22.2-29.6 GBq of 177-Lu DOTA-TATE: 3 (2%) obtained a CR, 32 (26%) a PR, 44 (35%) had SD and 22 (18%) developed PD. In a more extensive study by the same group the efficacy of ¹⁷⁷Lu-DOTA-TATE was evaluated in 310 patients and toxicity was evaluated in 510 patients, each receiving a cumulative radiation dose of 27.8-29.6 GBq in four treatment cycles with 6-10-week intervals between each cycle. Complete response was seen in 2 %, PR in 28% and a MR in 16% of patients. Acute side-effects such as nausea and vomiting, occurred after 25% and 10% of administrations, respectively. Sub-acute WHO haematological toxicity (grade 3 or 4) occurred in 3.6% of treatment cycles. Delayed toxicities included serious liver toxicity in two patients and myelodysplastic syndrome in 3 patients (53). Experience by Kwekkeboom et al. have led them to conclude that the two significant factors predicting favourable treatment outcome when using ¹⁷⁷LU-ocreotate were a high patient performance score and high uptake on the pre-treatment octreoscan (54).

In order to increase the efficacy of PRRT, an alternative option is to combine ⁹⁰Y-DOTA-TATE with ¹⁷⁷Lu-DOTA-TATE. In this way, the irradiation of the tumour mass is performed by a radioisotope with low-energy and short range (¹⁷⁷Lu) followed by a radioisotope with higher energy and wide range (⁹⁰Y). The results of a randomized study by Villard et al. on 486 patients with advanced GEP-NETs yielded an improvement of survival for patients treated with the combination (5.51 vs. 3.96 Gy) with comparable rates of toxicities (55).

In conclusion, radiolabelled somatostatin analogues have a good efficacy as therapeutic agents for PRRT in NETs (tumour reduction, improvement of quality of life, biochemical response). Only a small number of serious adverse effects occurred and only in those patients who had previous chemotherapy.