Phase I Study of Pasireotide (SOM 230) and Everolimus (RAD001) in Advanced Neuroendocrine Tumors

Abstract

Octreotide and everolimus have demonstrated efficacy in neuroendocrine tumors. Pasireotide is a somatostatin analog with binding affinity to a broader range of somatostatin receptor subtypes than octreotide. We performed a phase I study to evaluate the safety and feasibility of combining pasireotide with everolimus in patients with advanced neuroendocrine tumors. Cohorts of patients with advanced neuroendocrine tumors were treated with escalating doses of pasireotide (600–1200 mcg sc BID, followed by pasireotide LAR 40–60 mg IM monthly) and everolimus (5–10 mg daily). Twenty-one patients were treated. Dose limiting toxicities consisting of grade 3 rash and grade 3 diarrhea were observed. Twelve patients were safely treated at the maximum protocol-defined dose level of pasireotide LAR 60 mg IM monthly and everolimus 10 mg daily. Hyperglycemia was common; other observed toxicities were consistent with the known toxicities of either agent alone. Partial tumor response was observed in one patient; 17 (81%) patients experienced at least some tumor regression as their best response to therapy. In conclusion, pasireotide LAR 60 mg IM monthly in combination with everolimus 10 mg daily is feasible and associated with preliminary evidence of antitumor activity in patients with advanced neuroendocrine tumors. Further studies evaluating this combination are warranted.

INTRODUCTION

Somatostatin analogs have been widely used in patients with advanced neuroendocrine tumors for the treatment of carcinoid syndrome and related symptoms of hormone hypersecretion (di Bartolomeo, et al. 1996; Kvols, et al. 1986; Oberg, et al. 2004). More recently, treatment with the somatostatin analog octreotide has also been shown to slow tumor progression in patients with advanced carcinoid tumors. In the PROMID study, patients receiving octreotide had a significantly longer median time to tumor progression than those randomized to receive placebo (14.3 vs. 6 months) (Rinke, et al. 2009). The growth inhibitory effect of somatostatin analogs are thought to be mediated through binding to somatostatin receptors expressed on neuroendocrine tumor cells (de Herder, et al. 2003). Two biologically active forms of somatostatin, SS-14 and SS-28, are generated from cleavage of a pro-somatostatin peptide, and bind to five high-affinity G protein coupled somatostatin receptor subtypes (sst 1–5) (Lamberts 1991). Pasireotide (SOM230) is a multi-ligand somatostatin analog that has high binding affinity to the somatostatin receptors sst1, sst2, sst3, and sst5 (Schmid 2008). Compared with octreotide, pasireotide has greater binding affinity for sst1, sst3, and sst5 receptors and comparable affinity for sst2 (Schmid and Schoeffter 2004). The growth inhibitory effect of somatostatin and its analogs has been linked to direct activation of somatostatin receptors, as well as to indirect effects on growth factor production (Bevan 2005; Zatelli, et al. 2007). Binding to somatostatin receptors activates different protein tyrosine phosphatases that regulate intracellular effectors, including the extracellular signal-regulated kinase (ERK)1/2, phosphatidylinositol 3-kinase (PI3K)/AKT, and nitric oxide synthase (NOS) pathways, which leads to inhibition of cell proliferation and migration, or induction of apoptosis. Regulation of these pathways varies according to somatostatin subtype (Bousquet, et al. 2012; Oberg, et al. 2010). The increased binding to sst-1, sst-3, and sst-5 by pasireotide compared with octreotide may lead to additional antiproliferative activity and increased growth inhibition.

Pasireotide has been evaluated in vitro in human corticotroph or somatotroph tumor cell lines, where it was found to suppress cell proliferation (Batista, et al. 2006; Danila, et al. 2001). Pasireotide has also been associated with inhibition of cell proliferation in cell lines derived from small intestine or pancreatic neuroendocrine tumors (Kidd, et al. 2008). In clinical trials, pasireotide has shown preliminary evidence of efficacy in treating symptoms of hormone hypersecretion in patients with acromegaly or with octreotide-refractory carcinoid syndrome (Kvols, et al. 2006; Petersenn, et al. 2010). In these studies, doses of pasireotide ranged from 200–1200 micrograms (mcg) subcutaneously twice daily. Adverse events were generally mild, and included hyperglycemia, nausea, diarrhea, and abdominal pain. A long acting formulation, pasireotide LAR, administered monthly at doses of 40 or 60 mg resulted in drug exposure that was similar to that previously observed with the subcutaneous formulation at doses of 600 or 900 mcg administered twice daily (Wolin, et al. 2009). The side effect profile of pasireotide LAR was also similar to that observed with subcutaneous pasireotide.

The mTOR inhibitor everolimus is also associated with antitumor activity in patients with advanced neuroendocrine tumors. In initial phase II studies, treatment with everolimus was associated with partial responses and encouraging progression-free survival durations in both pancreatic neuroendocrine tumors and carcinoid tumors (Yao, et al. 2010; Yao, et al. 2008). In a subsequent randomized study in pancreatic neuroendocrine tumors, treatment with everolimus was associated with a significant improvement in progression free survival compared to placebo, leading to the approval of everolimus for this indication (Yao, et al. 2011). In a parallel placebo-controlled study performed in advanced carcinoid tumors, treatment with everolimus and octreotide was associated with a longer progression-free survival duration compared to octreotide alone, when measured according to local investigator assessment (Pavel, et al. 2011). These results suggest that everolimus is also associated with antitumor activity in carcinoid tumors; however, the study did not meet its primary statistical endpoint, which mandated improvement in progression-free survival based on central radiology review.

We performed a phase 1 study evaluating the safety and feasibility of combining pasireotide and everolimus in patients with pancreatic neuroendocrine or carcinoid tumors. Cohorts of patients were treated with escalating doses of pasireotide, beginning with the subcutaneous twice daily formulation and if tolerated, transitioned to the intramuscular LAR formulation. In parallel, everolimus was dose escalated from 5 mg to 10 mg in sequential cohorts. Patients were followed for evidence of toxicity and preliminary evidence of efficacy. Treatment was continued until tumor progression, unacceptable toxicity, or withdrawal of consent.

PATIENTS AND METHODS

Patient Population

All patients were required to be 18 years of age or older and have histologically documented, locally unresectable or metastatic carcinoid or pancreatic neuroendocrine tumors of low grade or intermediate histologic grade. Patients with poorly differentiated or high grade neuroendocrine carcinomas were not eligible. Treatment with prior chemotherapy was allowed, as was prior chemoembolization, cryotherapy, or radiofrequency ablation if measurable disease was not affected. Mandated laboratory requirements included: aspartate aminotransferase (AST) and alanine aminotransferase (ALT) ≤3 times the upper limit of normal (ULN) (<5 times ULN if liver metastasis was present), total bilirubin ≤2 times ULN, serum creatinine ≤1.5 times ULN, absolute neutrophil count (ANC) ≥1500/mm³, platelet count ≥100,000/mm3. All patients were required to have Eastern Cooperative Oncology Group (ECOG) performance status <2. Patients with uncontrolled diabetes mellitus or a fasting plasma glucose >1.5 times ULN were excluded. Patients with a history of congestive heart failure, myocardial infarction, or unstable angina pectoris within 6 months preceding enrollment were excluded. Patients with prolonged QTc (>450 msec at screening), a history of clinically significant cardiac arrhythmias, or on medications known to prolong QTc were excluded, as were patients with active or suspected chronic infections, including a history of chronic hepatitis B. All patients provided written informed consent for participation in the study, which was approved by the institutional review boards of participating institutions. The study was registered with clinicaltrials.gov (NCT00804336).

Treatment and Dose Escalation

Everolimus was administered orally as a once-daily dose. Pasireotide sc was self-administered subcutaneously twice daily for 4 weeks. If pasireotide sc was tolerated, patients received pasireotide LAR intramuscularly at the corresponding dose level. Pasireotide sc was continued for an additional 2 weeks after administration of pasireotide LAR until anticipated steady state levels of pasireotide LAR were achieved. Pasireotide LAR was administered every 28 days. Cycles for everolimus and pasireotide LAR were repeated every 28 days. The study schema is depicted in Figure 1. Everolimus and pasireotide were provided by the study sponsor (Novartis).

Figure 1.

Study schema.

• DLT was defined within the first 56 days of treatment.

• Dose levels are listed in Table 2.

A minimum of 3 and a maximum of 6 patients were enrolled in sequential cohorts according to a standard phase I dose-escalation design. Dose escalation to the next cohort proceeded in the absence of more than one of six patients experiencing a dose-limiting toxicity (DLT). DLT was defined as grade 3 or higher non-hematologic toxicity (excluding nausea, vomiting, hyperglycemia, hyperlipidemia, or alopecia), or grade 3 or higher hematologic toxicity lasting ≥ 7 days. DLT was defined within the first 56 days of treatment. Once the MTD was established, treatment of six additional patients at the MTD was planned to further characterize safety and toxicity.

Safety and Response Assessments

Patients were evaluated with a physical examination, serum chemistries, hematologic parameters, and for toxicity on day 1 and day 15 of the first 28-day treatment cycle, and on day 1 of subsequent treatment cycles. Additionally, all patients were required to monitor their fasting fingerstick blood glucose levels twice daily for the first week following initiation of study drugs, with additional monitoring recommended for patients who demonstrated evidence of hyperglycemia. For patients enrolled in the expanded cohort at the maximal dose level, ECGs to assess QTc were obtained at baseline, between 0.5 and 1.5 hours after the first pasireotide sc injection, on cycle 1 on day 15, and on day 1 and day 22 of treatment cycles 2–4. Toxicity was graded according to the National Cancer Institute (NCI) Common Terminology Criteria for Adverse Events (CTCAE), Version 3.0. Serum chromogranin and fasting lipid panel was obtained at baseline and after every two treatment cycles. Tumor response was evaluated at the end of every other 28-day treatment period using multi-phasic computed tomography (CT), or magnetic resonance imaging (MRI).

Pharmacokinetics

For patients in the expanded cohort treated at the maximum dose level, blood samples were collected for pharmacokinetic monitoring. Plasma trough concentrations of everolimus were determined by liquid chromatography-mass spectrometry on cycle 2 day 1 before dosing of everolimus. Plasma trough concentrations of pasireotide were determined using immunoassay at the following time points: cycle 2 day 1 (reflecting steady state concentration from pasireotide sc) and cycle 5 day 1 (reflecting steady state concentration from pasireotide LAR).

RESULTS

Patient characteristics

Characteristics of 22 patients enrolled in the study are listed in Table 1. The median age of the enrolled patients was 60 years, and 14 were male. Eighteen patients had carcinoid tumor, of whom most had small bowel primary tumors, and four patients had pancreatic neuroendocrine tumor. Eighteen had received prior treatment with octreotide; other prior systemic therapies included cytotoxic chemotherapy (8 patients) or biologically targeted therapies (13 patients). One patient had received prior treatment with everolimus, and one patient received prior temsirolimus.

Table 1.

Patient Characteristics

Characteristic	N (%)
Total patients enrolled	22
Median age (range)	60 (35–76)
Gender
Male	14 (64%)
Female	8 (36%)
ECOG performance status
0	16 (73%)
1 or 2	6 27%)
Tumor subtype
Carcinoid	18 (82%)
Small bowel	14
Bronchus	2
Unknown primary	2
Pancreatic neuroendocrine tumor	4 (18%)
Tumor histology
Well differentiated	17 (77%)
Moderately differentiated	5 (23%)
Prior therapies:
Octreotide	18
Cytotoxic chemotherapy	8
Biologically targeted therapy*	13
Interferon	4
Hepatic embolization	5
Progression within 6 months prior to study enrollment	20 (91%)

^*Prior biologically targeted therapies included: bevacizumab (12), sunitinib (4), everolimus (1), pazopanib (1), temsirolimus (1).

Dose escalation and Dose Limiting Toxicities

The dose escalation schema and associated DLTs are described in Table 2. No DLTs were observed in the first two patient cohorts. A single patient in cohort 3 developed grade 3 rash during the first 4 weeks of treatment. The cohort was expanded to a total of 6 patients with no further DLTs observed. Three patients were then enrolled in cohort 4, with one patient experiencing a DLT of grade 3 diarrhea following two weeks of treatment. After two weeks of treatment, another patient in cohort 4 developed grade 3 thrombocytopenia lasting <7 days requiring dose reduction. Although the protocol allowed enrollment of additional 3 patients to formally establish MTD, additional information regarding potential prolongation of the QTc interval associated with high dose pasireotide became available during the course of the study, and the investigators, together with the sponsor, elected to suspend further enrollment and treatment at this dose level. The three patients who started treatment at dose level 4 (everolimus 10 mg po qd and pasireotide 1200 mcg SC BID followed by pasireotide LAR 80 mg IM) were transitioned to dose level 3, and all received pasireotide LAR 60 mg IM rather than the previously planned 80 mg IM. Two patients received pasireotide 1200 mcg SC BID with everolimus 10 mg qd for two weeks before requiring dose reduction for toxicity, and one patient received treatment at this dose level for four weeks prior dose reduction due to the protocol modification.

Table 2.

Dosing Schema and DLT Summary

Dose level	Pasireotide SC BID (days 1–42)	Pasireotide LAR IM (day 29 and every 4 wks thereafter)	Everolimus (po qd)	N (21 evaluable)	DLT
1	600 mcg	40 mg	5 mg	3	None
2	900 mcg	60 mg	5 mg	3	None
3	900 mcg	60 mg	10 mg	6 +6 (expansion cohort)	Grade 3 rash (n=1)
4	1200 mcg	80 mg*	10 mg	3	Grade 3 diarrhea (n=1).

^*Dose level 4 was discontinued. Patients transitioned to dose level 3 prior to administration of pasireotide LAR 80mg.

An additional 6 patients were treated at dose level 3 (everolimus 10 mg po qd and pasireotide 900 mcg BID followed by pasireotide LAR 60 mg IM) in an expansion cohort, without observation of additional dose limiting toxicities.

Exposure to treatment and clinical toxicities

Twenty-one of the 22 enrolled patients received treatment and were evaluable for toxicity. Patients received a median of six 28-day treatment cycles; only 1 patient discontinued treatment due to toxicity. Other reasons for treatment discontinuation included disease progression (8) or physician/patient discretion (9). Three patients remained on study at the time of data analysis.

Suspected treatment-related adverse events across all treatment cycles were consistent with the expected toxicities of both everolimus and pasireotide. Treatment-related grade 3 and 4 adverse events included observations of grade 3 or 4 hyperglycemia in all 4 dose cohorts and in a total of 8 of the 21 evaluable patients (all 8 with carcinoid tumor). Hyperglycemia was managed with oral hypoglycemic agents alone in 2 patients; the remaining 6 patients were treated with insulin for initial management of their hyperglycemia. Of the 6 patients initially treated with insulin, 2 patients were subsequently able to transition to oral hypoglycemic agents. Following initial observations of hyperglycemia, the protocol was amended to include a requirement for daily blood glucose monitoring during the first treatment cycle and in subsequent treatment cycles if hyperglycemia persisted.

Other grade 3–4 non-hematologic toxicities were uncommon, and included one patient each with mucositis, rash, diarrhea, prolonged QTc interval (>500 ms) or joint pain. Grade 3–4 hypophosphatemia was observed in 6 patients. Grade 3–4 hematologic toxicites included thrombocytopenia (3 patients) and leucopenia (2 patients). Hypercholesterolemia, hypertriglyceridemia, diarrhea, and fatigue were common though mild (grade 1 or 2). We did not observe any statistically significant differences according to tumor type in the proportions of patients experiencing grade 3–4 toxicities or the most common overall toxicities. Grade 1–2 sinus bradycardia or QTc prolongation were each observed in 2 patients. In the expanded cohort of six patients treated at dose level 3, the mean baseline QTc interval prior to therapy was 429 msec. The mean maximal increase in QTc interval following treatment was 11 msec. In no cases were the cardiac anomalies associated with clinical sequelae.

Treatment efficacy

Patients were followed for biochemical and radiologic response with chromogranin A (CGA) levels and cross-sectional imaging studies after every 8 weeks of treatment. Nineteen patients had elevated CGA levels at baseline. Of these patients, 4 (21%) had a CGA level decrease of 50% or greater from baseline. Among 21 patients evaluable for radiologic response, 19 (90%) experienced stable disease as their best response to therapy, and 17 (81%) experienced some degree of tumor shrinkage during the course of treatment (Figure 2). We did not observe any differences in efficacy according to tumor type: all 4 patients with pancreatic neuroendocrine tumor and 16/17 patients with carcinoid tumor had a partial response or stable disease by RECIST criteria as their best response to treatment. One patient (5%) with a small bowel carcinoid tumor experienced a confirmed partial radiographic response; this patient also experienced a decrease in CGA level >50% from baseline. Two of the other patients with biochemical response experienced stable disease by RECIST associated with <10% decrease in tumor dimensions. The average time to best RECIST response was 3.6 months (range 1.6 to 12.8 months).

Figure 2.

Maximum reduction in sum of tumor diameters by patient and dose level.

* = progressive disease due to new lesions.

Of the treated patients, 91% had evidence of disease progression within 1 year prior to study enrollment. The proportion of patients on study who were progression-free at 6 months was 76%, and the proportion progression-free at 12 months was 65%.

Pharmocokinetics

For patients in the expanded cohort treated at dose level 3, trough concentrations of everolimus and pasireotide were determined at steady state. Based on its known half-life, a steady state trough concentration for everolimus was determined prior to dosing on cycle 2 day 1. The mean concentration of everolimus at this time point was 9.2 ng/ml (range 5.3 – 13.7). The mean trough concentration of pasireotide on cycle 2 day 1, reflective of a steady state from administration of pasireotide sc, was 7.9 ng/ml (range 4.5 – 11.9). The mean trough concentration of pasireotide on cycle 5 day 1, reflective of a steady state from administration of pasireotide LAR, was 18.0 ng/ml (range 11.5 – 26.1 ng/ml). Therapeutic levels of both agents were achieved.

DISCUSSION

This phase I study demonstrates the feasibility of combining the novel somatostatin analog pasireotide with everolimus in patients with advanced neuroendocrine tumors. Our results also show preliminary evidence of antitumor activity associated with this regimen. Patients remained on study for a median of nearly 6 months, and discontinuation due to adverse events was rare.

In general, the adverse events associated with this regimen were consistent with the anticipated toxicities of either agent alone. Hyperglycemia was one of the more commonly observed adverse events in our study and was observed in 90% of the treated patients; 6 (28%) patients required initiation of insulin. Hyperglycemia has been observed as an adverse event in previous studies of everolimus or pasireotide administered as single agents, albeit at lower prevalence. In the phase III registration study of everolimus for pancreatic neuroendocrine tumors, hyperglycemia was observed in 13% of patients in the everolimus treatment arm; in 5% the hyperglycemia was categorized as grade 3 or 4 (Yao et al. 2011). Hyperglycemia was reported in 6.7% of patients receiving pasireotide at doses of 200–600 mcg BID in a phase II study for the treatment of acromegaly (Petersenn et al. 2010). Hyperglycemia was also observed in 27% of patients receiving pasireotide at doses of 300–1200 mcg bid in a phase II study of pasireotide for the treatment of octreotide-refractory carcinoid syndrome (Kvols et al. 2006).

The higher prevalence of hyperglycemia in our study is likely related to the administration of both agents in combination, although the mechanisms underlying the development of hyperglycemia for these two agents remain poorly understood. The hyperglycemic effect of everolimus is an advantage in patients with insulinoma, where it appears to have a direct suppressive effect on insulin production and may also induce insulin resistance (Kulke, et al. 2009). The hyperglycemic effect of somatostatin analogs are thought to be caused in part through suppression of insulin secretion, mediated by binding to somatostatin receptor subtypes (sst) 1, 2, and 5 (Singh, et al. 2007). Pasireotide has a higher affinity for sst5 than octreotide, and may contribute to the hyperglycemia observed with this agent. The mechanism underlying the hyperglycemic effect of pasireotide may also include indirect suppression of GLP-1, resulting in stimulation of glucagon secretion (Holst, et al. 2011). Close monitoring and appropriate treatment of hyperglycemia in future studies of this regimen are warranted.

Cardiac side effects of somatostatin analogs have been previously reported; the most common side effect associated with octreotide is bradycardia (Herrington, et al. 1998). Asymptomatic bradycardia was observed in 2 patients receiving pasireotide in our study. Concerns regarding the potential for high doses of pasireotide to prolong the QTc interval were also raised while our study was in progress. In a study involving healthy adults designed to examine cardiac safety of pasireotide, a dose of 1950 mcg twice daily demonstrated a possible QTc prolongation effect (Novartis, 2012). This concern resulted in the discontinuation of dose level 4. As a result, no patients were treated with pasireotide LAR 80 mg, and a formal MTD was not defined. While prolonged QTc was observed in 4 patients in our study, in no case was the QTc prolongation associated with confirmed clinical sequelae.

A variety of options exist for the management of advanced neuroendocrine tumors, including surgical, medical, and nuclear medicine strategies (Boudreaux, et al. 2010; Kulke, et al. 2010; Pavel, et al. 2012). Both everolimus and somatostatin analogs have been associated with antitumor activity in advanced neuroendocrine tumors (Pavel et al. 2011; Rinke et al. 2009; Yao et al. 2010; Yao et al. 2008). Combining everolimus with somatostatin analogs offers a promising treatment approach due to potential synergistic effects on the PI3K-mTOR pathway (Bousquet et al. 2012). In our study, the combination of pasireotide and everolimus was associated with tumor regression in 17/21 (81%) of patients, and with a RECIST-defined partial response in one patient. This value compares favorably to a 64% rate of regression associated with everolimus alone in a randomized trial of pancreatic neuroendocrine tumors (Yao et al. 2011). While comparisons between these studies are clearly limited by the small size of our study as well as our inclusion of patients with either pancreatic neuroendocrine tumors or carcinoid tumors, our results support further studies exploring the antitumor effect of this combination. If future studies confirm efficacy of everolimus together with pasireotide, this combination could represent an important treatment option for patients.

The recommended dose for further studies based on our results is pasireotide LAR 60 mg monthly, administered in combination with everolimus 10 mg daily, which represents the dose currently approved for the treatment of pancreatic neuroendocrine tumors. We note, however, that dose escalation in our study was halted prior to formal determination of the MTD. It is possible that pasireotide LAR at higher doses may have greater efficacy, and further phase I dose escalation studies of pasireotide LAR, with appropriate monitoring, are warranted.

Table 3.

Number of Patients Experiencing Selected Adverse Events by Dose Level

	Dose level 1 (n=3)		Dose level 2 (n=3)		Dose level 3 (n=12)		Dose level 4 (n=3)
AE Grade	1–2	3–4	1–2	3–4	1–2	3–4	1–2	3–4
Non-hematologic AEs
Hyperglycemia	2	1	2	1	7	4		2
Hypercholesterolemia	3				7		1
Hypertriglyceridemia	2		2		5		1
Hypomagnesemia			3		5
Hypophosphatemia		1		1	3	3	1	1
Hypocalcemia			2		6		2
Alkaline Phosphatase			1	1	5	1	1
AST	1		1		7		2
Fatigue	2		2		9		3
Diarrhea	1		1		8		2	1
Nausea	2		2		9		2
Mucositis	1		1		6	1	1
Rash	1		1		6	1	1
Joint pain								1
QTc prolongation					2	1	1
Sinus bradycardia			1				1
Hematologic AEs
Leukocytopenia	1		2		7		2
Lymphopenia		1	1			1
Thrombocytopenia	1		2		9	2	1	1
Anemia	2		3		8		3