Advances in the Treatment of Pancreatic Neuroendocrine Tumors (pNETs)

ABSTRACT

Recent clinical trials have led to significant advancements in treatment options for metastatic neuroendocrine tumors of the pancreas. Sunitinib and everolimus have been approved by the Food and Drug Administration for treatment of progressive pancreatic NETs based on phase III trial data demonstrating improvements in progression-free survival. Cytotoxic drugs such as temozolomide and capecitabine have been associated with high radiographic response rates; however data derives primarily from subset analysis of prospective trials and from retrospective series. During the next few years, randomized clinical trials are expected to provide more clarity on the role of somatostatin analogs and cytotoxic drugs. New studies are also evaluating biomarkers that will potentially allow for improved selection of drugs for specific tumor subtypes.

The past several years have seen considerable advances in the management of pancreatic neuroendocrine tumors (pNETs). New staging and grading classifications have aided in the prognostic stratification of these rare tumors. Even more importantly, insights into the biology of pNETs have led to development and approval of new treatments. In this article, we review the recent advances in systemic and liver-directed therapies of metastatic pNETs.

SOMATOSTATIN ANALOGS

The somatostatin analogs (SSAs) octreotide and lanreotide were initially developed to palliate hormonal symptoms associated with metastatic neuroendocrine tumors.¹ Hormonally active pNETs that secrete vasoactive intestinal peptide and glucagon appear to be particularly sensitive to the inhibitory effects of SSAs on hormone-related symptoms.² In recent years, accumulating data have also supported the role of SSAs as antiproliferative agents capable of stabilizing tumor growth in patients with metastatic NETs.³ Although the precise mechanisms are not fully understood, the inhibitory effects on tumor growth appear to involve activation of intracellular phophotyrosine phosphatases ⁴ and suppression of circulating cytokines such as insulin-like growth factor.⁵ The PROMID trial, a randomized study comparing octreotide LAR to placebo, demonstrated a significant improvement in time for tumor progression among patients with advanced small intestinal (midgut) NETs.⁶ It remains unclear whether SSAs exert a similar inhibitory effect on growth of pNETs. A randomized trial evaluating lanreotide vs. placebo in nonfunctioning gastroenteropancreatic NETs is ongoing; results are expected in 2013.

EVEROLIMUS

The mammalian target of rapamycin (mTOR) is a serine/threonine kinase that regulates cell metabolism and proliferation in response to environmental stimuli.⁷ The mTOR enzyme lies downstream of the PI3K/AKT pathway and is upregulated in a variety of malignancies in response to stimulation by growth factors and cytokines. Recent whole-exome genomic analysis of pNETs demonstrates that approximately 15% of tumors are associated with somatic mutations in genes associated with the mTOR pathway such as PTEN, PI3K, and TSC2.⁸

The oral mTOR inhibitor everolimus was studied in a large phase II trial of 160 patients with advanced, progressive PNETs.⁹ Patients were evaluated in two strata: everolimus monotherapy (n = 115) or everolimus plus octreotide (n = 45). Response rates and median progression-free survival (PFS) were 9% and 9.7 months, respectively, in the monotherapy arm vs. 4% and 16.7 months in the combined therapy arm. In the phase III RADIANT 3 trial, 410 patients with advanced, progressive pNETs were randomized to everolimus 10 mg once daily vs. placebo, with unblinding and crossover allowed upon disease progression. The trial demonstrated a clinically and statistically significant improvement in PFS from 4.6 months on placebo to 11 months on everolimus (HR 0.35; p = <0.0001), thus meeting its primary end point.¹⁰ No differences were observed in overall survival (OS), possibly due to the trial design, which allowed nearly all patients receiving placebo to cross over to the treatment arm. Based on the results of this study, everolimus received an indication from the Food and Drug Administration for unresectable or metastatic pNETs. Side effects of everolimus include aphthous oral ulcers, rash, hyperglycemia, interstitial pneumonitis, and immunosuppresion.¹⁰

SUNITINIB

Neuroendocrine tumors are highly vascular and frequently overexpress the vascular endothelial growth factor (VEGF) ligand and receptor (VEGFR).¹¹ The tyrosine kinase receptor inhibitor sunitinib targets VEGFR-1, -2, and -3, as well as platelet-derived growth factor receptor, KIT, and FLT3. In a phase II study of metastatic NETs, an objective response rate of 16.7% was observed among the cohort of patients with pNETs.¹² Based on these promising data, a phase III study was launched comparing sunitinib 37.5 mg daily vs. placebo in patients with advanced, progressive well-differentiated pNETs. The study was initially designed to enroll 340 patients; however, it was stopped at interim analysis, after enrollment of 171 patients, due to a significant improvement in median PFS from 5.5 months on the placebo arm to 11.4 months on the treatment arm (HR 0.418; p = .0001).¹³ Although the trial did not have a built in crossover, unblinding of patients at the time of interim analysis allowed the majority of placebo patients to cross over to the treatment arm. A trend toward improved OS was observed, which did not meet statistical significance. Side effects of sunitinib include hypertension, palmar-plantar erythrodysesthesia, fatigue, diarrhea, and cytopenias.¹³

CYTOTOXIC CHEMOTHERAPY

Although NETs of the digestive tract (carcinoid tumors) are often considered chemoresistant, accumulating evidence suggests that pNETs are considerably more sensitive to the effects of cytotoxic drugs. Streptozocin (STZ)-based combinations have long been effective in management of advanced pNETs, particularly among patients with relatively rapid tumor growth or high tumor burden. Two randomized trials conducted in the 1970s and 1980 reported response rates of 63% with STZ plus 5-fluorouracil (5-FU)¹⁴ and 69% with the combination of STZ and doxorubicin.¹⁵ The reproducibility of these results has subsequently been challenged due to the partial reliance on nonradiographic response criteria.¹⁶ However, a more recent retrospective study investigating the combination of STZ, 5-FU, and doxorubicin reported a response rate of 39% using objective radiographic criteria.¹⁷ Another alkylating agent, dacarbazine, was associated with an objective response rate of 33% in a phase II trial of 42 patients.¹⁸

The clinical use of STZ and dacarbazine has been limited by toxicity concerns. In recent years, the oral alkylating agent temozolomide has emerged as an active agent in pNETs. Like dacarbazine, temozolomide is converted to the active alkylator MTIC, which induces DNA methylation. A phase II study investigating the combination of temozolomide and thalidomide demonstrated an objective response rate of 45% in the subset of 11 patients with pNETs.¹⁹ A subsequent retrospective study of temozolomide combined with capecitabine in 30 chemo-naive patients reported an objective radiographic response rate of 70% and a median PFS of 18 months.²⁰ A recently published phase II trial of temozolomide and bevacizumab reported a response rate of 33% in a more heavily pretreated subset of pNETs.²¹ An upcoming randomized phase II clinical trial sponsored by the Eastern Cooperative Oncology Group (ECOG) is designed to compare temozolomide monotherapy vs. temozolomide plus capecitabine. It will also investigate whether expression of the DNA repair enzyme MGMT predicts response to chemotherapy.

LIVER-TARGETED THERAPIES

The liver is the predominant site of metastases in patients with advanced pNETs. Hepatic directed therapies include surgical resection, radiofrequency ablation, and transarterial embolization or chemoembolization. A novel approach to liver metastases involves arterial embolization of the beta-emitting radioactive isotope yttrium-90 (⁹⁰Y) embedded in either resin microspheres (SIR-sphere) or glass microspheres (TheraSphere). This technique enables delivery of radiation directly to liver tumors. In one retrospective multicenter study of 148 NET patients treated with SIR-Spheres, the objective radiographic response rate was 63%.²² Short-term toxicities associated with ⁹⁰Y radioembolization appear to be lower than with other embolization techniques; however, there are concerns about long-term toxicities including the effects of radiation hepatitis in patients with large volume disease. In the absence of randomized trials, it is quite difficult to recommend a particular embolization modality. The combination of transarterial embolization with antiangiogenic treatment has also been of interest. In one phase II trial of 39 patients with advanced NETs treated with embolization followed by sunitinib, a response rate of 72% was observed with a median PFS of 15 months.²³

INTEGRATING THE DATA

The recent advances highlighted in this review have expanded the therapeutic armamentarium available to physicians who treat metastatic pNETs. However, new questions are raised: Who should be treated, how should treatments be sequenced, and how should they be personalized for the individual patient? Evidence supporting use of the biological agents everolimus and sunitinib includes randomized, placebo-controlled studies demonstrating statistically significant improvements in PFS. However, response rates associated with these agents are quite low. Temozolomide- and streptozocin-based regimens appear to yield substantially higher response rates; however, temozolomide-based regimens have not been studied in randomized clinical trials. At this time, no proven predictive factors can help guide treatment selection. Indeed, guidelines published by the National Comprehensive Cancer Network (NCCN) list somatostatin analogs, everolimus, sunitinib, cytotoxic chemotherapy, and liver-targeted therapies as treatment options; however, the guidelines do not recommend a particular sequence of treatments.

In our practice, we typically use somatostatin analogs as first-line therapy for patients with low-grade tumors that express somatostatin receptors (as evidenced on OctreoScan). Hormonally functioning tumors, such as VIPomas or glucagonomas, should also be treated with somatostatin analogs in the first line for symptomatic control. We use cytotoxic agents, such as capecitabine/temozolomide, in patients with high-tumor burden, relatively high proliferative activity (mitotic rate or Ki-67 index), relatively rapid tumor progression, or high symptomatic burden. The choice between everolimus vs. sunitinib is difficult given the very similar improvements in PFS observed on phase III clinical trials. Selecting between these two agents often involves a discussion of competing side effects.

CONCLUSIONS

Given the relatively long survival durations observed among patients with advanced pNETs,²⁴ multiple lines of therapy can often be given. Current options include SSAs, everolimus, sunitinib, temozolomide- or streptozocin-based cytotoxic regimens, and liver-targeted therapies. It is important to avoid overtreatment of asymptomatic patients with stable or mildly progressive disease since long-term treatment can adversely impact quality of life. Future clinical trials should address the selection and sequencing of treatments in order to balance efficacy with toxicity.