MEN1 syndrome represents a difficult challenge for surgeons mainly because it is associated with heterochronous and multifocal NF-PNETs. Prophylactic pancreatectomies could remove all premalignant lesions prior to malignant conversion but are not an accepted approach. Therefore, the careful monitoring of these patients for the early diagnosis and surgical excision of any arising tumors with malignant potential is of utmost importance [55]. Based on previous evidence, the European Neuroendocrine Tumor Society (ENETS) recommends operating MEN1-related NF-PNETs with a diameter of more than 2 cm, with a yearly size increment in diameter of more than 0.5 cm, or with metastases [30, 105]. The management of NF-PNETs of less than 2 cm is debated as they may also have an aggressive biology, but they usually are microadenomas and have a more indolent behavior, and their management is close monitoring.

Overall, the refinement and personalization of surgical procedures in NF-PNETs have resulted in improved long-term survival in patients with locally advanced and metastatic disease, as well as shorter hospital stays and comparable long-term outcomes in patients with limited disease treated minimally invasively. There are still controversies related to issues of surgical treatment of PNETs, such as to what extent enucleation, lymph node sampling and vascular reconstruction are beneficial for the oncologic outcome particularly in MEN1 patients and in tumors 1–2 cm in diameter. It is expected that the recently developed endoscopic ultrasound-guided biopsies will shed light to all these gray zones.

The clinical application of long-acting somatostatin analogs (SSAs) has changed dramatically the quality of life and prognosis of patients with NETs [79]. Their benefit was initially thought to be related to their antisecretory effect as they were shown to relieve patients from troubling clinical syndromes related to the peptide secretion. Altogether, SSAs have been shown to be well tolerated and safe in all available forms as patients may experience only mild and infrequent side effects such as abdominal pain, steatorrhea, and cholelithiasis [81]. This last effect is due to the inhibitory effects of octreotide on gallbladder contractility, and although it requires follow-up attention, is rarely of clinical significance.

From the advent of the first SSA, an improved survival has been shown in NET patients, and although a clear gain could be seen by the fact that patients’ quality of life was improved together with their symptoms, it was not clear whether there was also an antiproliferative effect. Several groups have reported a possible antiproliferative action from SSA, by induction of apoptosis [1, 23, 44], although significant tumor shrinkage has been reported to occur in <5 % of cases [79]. Initial in vitro studies suggested a number of direct and indirect mechanisms of such an effect. Direct mechanisms involve the activation of somatostatin receptors (SSRs) leading to modulation of intracellular signaling transduction such as the activation of MAP kinase pathway [118]. Indirect antiproliferative effects are exerted through activation of SSRs on endothelial cells and monocytes leading to blocking of tumor angiogenesis [20, 49]. A proof of direct antiproliferative effect came from the PROMID trial, a phase III study that randomized patients to receive octreotide long-acting repeatable (LAR) or placebo [94]. This placebo-controlled, double-blind study involved patients with well-differentiated metastatic NETs originating in the distal intestine and proximal colon and showed a clear benefit in all patients receiving octreotide including those with NF tumors. Nevertheless, in this trial, disease stabilization was the main clinical response observed as hardly any tumor shrinkage was achieved and there were no patients with NF-PNETs included in this study. Furthermore, no clear prospective evidence supporting the antitumoral effect of SSAs in NF-PNETs exists. However, based on these encouraging results, octreotide LAR was recommended [84] and approved in several countries in progressing well-differentiated, metastatic nonfunctioning NETs, including PNETs. Supporting are the results of a randomized phase III double-blind trial testing lanreotide versus placebo in NF well-differentiated enteropancreatic NETs (the CLARINET study; clinical trial link: http://​clinicaltrials.​gov/​show/​NCT00353496). In a previous study including 21 well-differentiated advanced NF-PNETs, treatment with octreotide LAR was associated with disease stabilization and good quality of life in 38 % of patients [10]. A Ki-67 > or =5 % and/or weight loss was associated with more aggressive disease and would justify more aggressive therapy [10].

Early studies had shown that interferon-alpha (IFN-α) has antiproliferative, through Jak1/Tyk2 kinase and STAT1/STAT2 activation, antisecretory, antiangiogenic, and antiapoptotic effects [22, 76, 109]. A possible positive immunomodulatory effect on T-cell and natural killer activity was also suggested [39, 77], but it has not been confirmed since by later studies. Interferon-α may be used as monotherapy or in combination with SSAs, as antisecretory and antiproliferative treatment, with variable response rates (RRs). There has been biochemical response in 40–60 % of patients, symptomatic improvement in 40–70 % of patients, and significant tumor shrinkage in a median of 10–15 % of patients [37, 78, 82]. Side effects of IFN-α are significant and include flu-like syndrome, chronic fatigue, depression, weight loss, polyneuropathy, myositis, thrombocytopenia, anemia, leukopenia, and autoantibodies in 15 % of patients (autoimmune hepatitis/thyroiditis, psoriasis, systemic lupus erythematosus syndrome with antinuclear factors, and parietal cell antibodies with pernicious anemia) [78, 84].

One of the first reports to compare results from different studies on octreotide and on IFN-α monotherapy in NETs suggested that better biochemical responses and disease stabilization were obtained with octreotide and not with IFN-α [17]. An initial single-arm study suggested that for patients with metastasized endocrine gastroenteropancreatic tumors, the combination of octreotide and IFN-α is superior to either as monotherapy [75]. Another early study showed that the addition of IFN-α to octreotide has antiproliferative efficacy in a subgroup of patients with advanced metastatic disease unresponsive to octreotide monotherapy [37]. Similarly, in another study, patients having progressed on either octreotide or IFN-α benefited from the subsequent combination of both drugs [36]. There are very few studies comparing SSA monotherapy to IFN/SSA combinations, and the results are conflicting. There are no studies employing IFN monotherapy or in combination with SSA in NF-PNETs. A prospective randomized study showed a significantly reduced risk of tumor progression during follow-up in patients with midgut carcinoid tumors metastatic to the liver receiving octreotide and IFN-α as compared to single-agent octreotide [50]. A randomized phase III study comprised three arms to compare lanreotide, IFN and lanreotide plus IFN [28]. All arms had equivalent results with respect to symptom control. No arm showed statistically significant survival improvement, but there was a trend of better survival in the combination arm. In the same study, the most toxic arm was the combination one. Another recent randomized trial comparing octreotide to octreotide/IFN-α showed no survival advantage between the two arms [4]. It is hypothesized that the difference in the results lies on the heterogeneity of the disease and further larger trials are required. At present, we do not have enough statistical evidence for an upfront use of the combination of IFN-α and SSA in patients with NETs, but we have some clinical evidence coming from non-randomized studies and the sub-analysis of randomized trials that would justify the sequential use of the two drugs or the combination after progression to single-agent therapy [33].

Head to head comparisons of octreotide/IFN-α to other biological combinations have not been performed yet. A randomized phase III trial is currently recruiting patients in order to compare the efficacy of octreotide/IFN-α2b and octreotide/bevacizumab in metastatic or locally advanced, high-risk NETs (clinical trial link: http://​clinicaltrials.​gov/​show/​NCT00569127). This trial was based on a previous phase II study showing better PFS and objective responses of octreotide/bevacizumab compared to octreotide/PEG-IFN [116].

The more thorough understanding of signaling pathways implicated in tumor cell proliferation and angiogenesis underlining the pathogenesis of NETs has prompted the application of small molecules targeting these pathways such as everolimus and sunitinib.

Over the last three decades, the applied cytotoxic agents for NETs included DNA-damaging agents such as the antimetabolite 5-fluorouracil (5-FU); the alkylating agents streptozotocin (STZ), dacarbazine (DTIC), and temozolomide (TMZ); the topoisomerase inhibitors doxorubicin (Dox) and etoposide; and the DNA crosslinker platinum derivatives, such as cisplatin (CDDP), carboplatin (CBDCA), and oxaliplatin. Single-agent chemotherapy has not shown any real benefit [69], and the newer chemotherapy regimens based on the combination of these active drugs show low RRs, which set the need to improve the results of the medical treatment of these malignancies [108].

Chemotherapy has heterogeneous results in NET patients mainly due to the large variability of biological features among the disease subgroups. For example, chemotherapy in functioning advanced NETs has limited efficacy with a RR of 15 % [102], whereas PNETs have a RR of 30–40 % although with significant toxicity [26, 67]. One of the earliest observations regarding these tumors was that differentiation grade represents one of the most important predictive markers of tumor response to chemotherapy. Patients with metastatic disease of well- and moderately differentiated NF-PNETs are expected to respond better to STZ-based and targeted therapy, whereas those of high-grade NF-PNETs with Ki-67 LI above 20 % are likely to benefit from platinum-based treatment. More specifically, an early report from the Mayo Clinic suggested that anaplastic NETs independently of the primary site location are responsive to combination chemotherapy including etoposide and cisplatin [70]. Conversely, well-differentiated tumors were unresponsive to this regimen [70]. Overall, the population with anaplastic disease had an overall RR of 67 %, a median time to progression (TTP) of 8 months and a median survival of 19 months. Although toxicity of this regimen was severe with vomiting, leukopenia, thrombocytopenia, anemia, alopecia, and neuropathy, the efficacy of this regimen remains unsurpassed by any other treatment until today as there is no unequivocal evidence of survival improvement. Similarly to the grade of differentiation, recent evidence suggests that a careful evaluation of the mitotic index and the Ki-67 assessment can provide further information on the overall prognosis and the degree of response to systemic cytotoxic treatment. For instance, highly proliferative (Ki-67 >30 %) tumors with aggressive behavior such as the PNETs are more sensitive to cytotoxic therapy with cisplatin/etoposide [5]. On the other hand, tumors originating from the small intestine have most commonly a low mitotic index and demonstrate poor response to cytotoxic chemotherapy [5].

One of the first tested combinations was STZ/5-FU in the 1970s and remained a standard treatment until now [68]. There have been very few phase III trials in NETs with conflicting results mainly due to the heterogeneity of the population taken in consideration. The first comparing 5-FU/STZ to Dox showed no statistically significant difference between the two arms [27], and the second comparing 5-FU/Dox to 5-FU/STZ showed the latter being statistically significant better [102]. For PNETs, there have been two phase III trials, one comparing STZ to STZ/5-FU that found no difference between the two arms [69] and a second one comparing chlorozotocin, STZ/5-FU, and STZ/Dox showing a statistically significant benefit in the last arm [71]. During the past decade, there have been few chemotherapy studies in advanced well- and moderately differentiated NETs with encouraging results: (a) with STZ and Dox [21]; (b) with STZ, 5-FU, and Dox [54]; (c) with 5-FU, epirubicin, and dacarbazine [110]; and (d) with STZ, 5-FU, and cisplatin [106]. For the poorly differentiated NETs, the most effective combination chemotherapy remains cisplatin with etoposide [45]. Interestingly, it has been recently reported that gastroenteropancreatic carcinomas with Ki-67 <55 % had a lower RR to platinum-based chemotherapy while having a better prognosis compared with those having Ki-67 >55 % [100]. As a result, it is expected that patients with Ki-67 <55 % may respond to different chemotherapy schedules such as temozolomide, but this needs to be further investigated in the context of a clinical trial [53].

The very high response rates reported in the first studies with STZ-based chemotherapy were criticized since they derived in part from the historical use of nonstandard response criteria. Furthermore, STZ-related toxicity and not manageable schedule limited its use and prompted to newer and better manageable agents like TMZ.

An interesting novel combination chemotherapy is an oral regimen with capecitabine and TMZ that has been shown to exert high activity and low toxicity both in well-differentiated [101] and poorly differentiated neuroendocrine tumors [112]. However, a larger study comparing this regimen with STZ-based therapy or TMZ monotherapy is missing.

Based on the findings of a study assessing retrospectively the triplet capecitabine, STZ, and CDDP, the National Cancer Research Network approved NET01, a randomized phase II study comparing two chemotherapy regimens, capecitabine/STZ and capecitabine/STZ/cisplatin, in unresectable metastatic NETs of the foregut and pancreas and NETs of unknown primary site [106].

Another interesting approach with promising efficacy is the combination of targeted agents with chemotherapy. One such doublet is TMZ and the antiangiogenic agent bevacizumab that has been explored in two different TMZ schedules, a standard [15] and a low continuous dose [52]. A second doublet is TMZ with everolimus that has been tested in 40 patients with PNETs and is shown to achieve a PR in 40 % and a PFS of 15.4 months [13]. However, further studies evaluating the efficacy of these combination therapies compared to TMZ alone are warranted.