Both everolimus and temozolomide are associated with single-agent activity in patients with pancreatic neuroendocrine tumor (NET). A phase 1/2 study was performed to evaluate the safety and efficacy of temozolomide in combination with everolimus in patients who have advanced pancreatic NET.
Patients were treated with temozolomide at a dose of 150 mg/m2 per day on days 1 through 7 and days 15 through 21 in combination with everolimus daily in each 28-day cycle. In cohort 1, temozolomide was administered together with everolimus at 5 mg daily. Following demonstration of safety in this cohort, subsequent patients in cohort 2 were treated with temozolomide plus everolimus at 10 mg daily. The duration of temozolomide treatment was limited to 6 months. Patients were followed for toxicity, radiologic and biochemical response, and survival.
A total of 43 patients were enrolled, including 7 in cohort 1 and 36 in cohort 2. Treatment was associated with known toxicities of each drug; no synergistic toxicities were observed. Among 40 evaluable patients, 16 (40%) experienced a partial response. The median progression-free survival duration was 15.4 months. Median overall survival was not reached.
With the recent approval of the targeted agents everolimus and sunitinib, a broad range of systemic therapeutic options can now be considered for the treatment of advanced pancreatic neuroendocrine tumor (NET).[1, 2] In this study, we explored the antitumor activity of everolimus administered in combination with temozolomide, an oral alkylating agent that has been associated with antitumor activity in advanced pancreatic NET.
Single-agent activity of the mammalian target of rapamycin (mTOR) inhibitor everolimus has been observed in phase 2 studies and subsequently confirmed in a randomized phase 3 study of patients with pancreatic NET.[1, 3, 4] Although everolimus has been associated with a clear improvement in progression-free survival duration compared with placebo (11 versus 5.5 months), the overall tumor response rate is low (4%). Similarly, a low objective response rate (9%) has been observed in patients with pancreatic NET treated with the tyrosine kinase inhibitor sunitinib.
Because of the low response rates associated with these targeted agents, patients with a high tumor burden or symptoms from bulk of disease may benefit from treatment with a cytotoxic agent due to their higher response rates. Although both streptozocin and dacarbazine have been associated with antitumor activity in pancreatic NET,[5, 6] the use of these intravenously administered drugs for advanced pancreatic NET has not been universally adopted. In contrast, temozolomide, an oral analogue of dacarbazine, has been increasingly used for the treatment of pancreatic NET. Retrospective studies evaluating single-agent temozolomide in this setting have reported overall response rates of 8% to 25%.[8, 9] In a larger series of 53 patients with pancreatic NET treated with various temozolomide-based regimens, 18 (34%) experienced partial or complete responses. The highest activity reported to date of temozolomide-based therapy is from a recent single-institution retrospective study in which 30 patients with advanced pancreatic NET were treated with temozolomide in combination with capecitabine. In this study, the reported response rate was 70%, and the median progression-free survival duration was 18 months.
Prospective studies of temozolomide-based regimens appear to confirm the activity of temozolomide in patients with advanced pancreatic NET. In 2 single-institution phase 2 studies, temozolomide has been administered using a dose-intense schedule of 150 mg/m2 per day administered for 7 days every other week.[12, 13] In the first study, which assessed the combination of temozolomide and thalidomide, 5 of 11 patients with pancreatic NET experienced partial responses. In a follow-up study evaluating the combination of temozolomide and bevacizumab, objective tumor responses were observed in 5 of 15 (33%) patients; the median progression-free survival duration for patients with pancreatic NET in this study was 14.3 months. In another recent study of patients with NET from various sites of origin, temozolomide was administered at a dose of 100 mg daily in combination with bevacizumab at 7.5 mg/kg every 3 weeks. Responses were observed in 9 of 15 (60%) patients.
In light of the demonstrated activity of both everolimus and temozolomide in patients with advanced pancreatic NET, we sought to explore the safety and antitumor activity of these 2 agents administered in combination. We performed a prospective, phase 1/2 study at 3 institutions. Temozolomide was administered using a dose-intense regimen of 150 mg/m2 every other week. Everolimus was administered in combination with temozolomide at a dose of 5 mg in an initial safety cohort, and following evidence of tolerability, at a dose of 10 mg daily in a second cohort. Treatment with temozolomide was limited to 6 months to limit long-term toxicity, and prophylaxis with trimethoprim-sulfamethoxazole was required.
MATERIALS AND METHODS
Patients were required to have histologically confirmed low- or intermediate-grade (grade 1 and 2) metastatic or locally unresectable pancreatic NET. Patients with poorly differentiated or high-grade neuroendocrine carcinoma were excluded. Patients were further required to have measurable disease by Response Evaluation Criteria in Solid Tumors (RECIST) criteria; Eastern Cooperative Oncology Group (ECOG) performance status of ≤ 2; adequate hepatic function (serum bilirubin ≤ 1.5 times upper limit of the normal range [ULN]); aspartate aminotransferase ≤ 3 times ULN (≤ 5 times if liver metastases were present); and adequate bone marrow function (absolute neutrophil count ≥ 1500/mm3; platelets ≥ 100,000/mm3). Patients with fasting serum cholesterol > 300 mg/dL or fasting triglycerides > 2.5 times ULN could begin treatment after initiation of appropriate lipid lowering medication. Patients were excluded if they had received prior treatment with an mTOR inhibitor or temozolomide or if they were receiving chronic treatment with systemic steroids or other immunosuppressive agents.
Patients were enrolled in the study between June 2008 and November 2010. All patients provided signed, informed consent as required by the institutional review boards of the participating centers, which included the Dana-Farber Cancer Institute, Massachusetts General Hospital (both in Boston, Mass), and Lahey Clinic Medical Center (Burlington, Mass).
Temozolomide was administered to all patients at a starting dose of 150 mg/m2 on days 1 to 7 and days 15 to 21 of a 28-day cycle. Everolimus was administered together with temozolomide in 2 dose cohorts. Cohort 1 was designed to consist of a minimum of 3 evaluable patients treated with everolimus at a dose of 5 mg daily. If a dose-limiting toxicity (DLT) was observed in these patients during the first 28 days, 3 additional evaluable patients were to be enrolled with dose escalation to cohort 2 only if no additional DLTs were observed. In cohort 2, patients received everolimus 10 mg daily together with temozolomide. DLT was defined as grade ≥ 3 nonhematologic toxicity, grade ≥ 2 neutropenia or thrombocytopenia failing to revert to ≤ grade 1 within 2 weeks, febrile neutropenia, or inability of a patient to take > 75% of the planned dose during the treatment period.
Because of the risk for temozolomide-related selective lymphopenia, the maximum treatment duration of temozolomide was limited to 6 months. Patients also received prophylaxis against Pneumocystis jirovecii (formerly P. carinii) with trimethoprim-sulfamethoxazole, 1 tablet given orally every Monday, Wednesday, and Friday. Patients with allergies to trimethoprim-sulfamethoxazole received an alternate, standard pneumocystis prophylaxis regimen. After completing 6 months of therapy with temozolomide, everolimus alone was continued until disease progression, withdrawal of consent, or unacceptable toxicity.
Dose adjustments for temozolomide and everolimus were allowed for both hematologic and nonhematologic toxicity. Treatment was held if patients developed an absolute neutrophil count < 1000/mm3 or a platelet count < 50,000/mm3 and was not resumed until full hematologic recovery. On recovery, treatment was resumed with a dose reduction of temozolomide by 50 mg/m2 and dose reduction of everolimus to 5 mg daily if receiving 10 mg daily or to 5 mg every other day if receiving 5 mg daily. Treatment with both drugs was also held for grade ≥ 2 stomatitis or pneumonitis or other grade ≥ 3 nonhematologic toxicity. Treatment was resumed with dose reduction pending resolution of nonhematologic toxicities to grade ≤ 1.
Radiologic tumor assessment and biochemical assessment with plasma chromogranin A levels were performed every 2 cycles (8 weeks) after initiation of treatment.
The primary objective of this study was to determine the radiographic response rate for the combination of temozolomide and everolimus in patients with advanced pancreatic NETs. Secondary objectives included assessment of progression-free survival (PFS), overall survival (OS), biochemical response rate, and toxicity.
Responses were determined by RECIST criteria. Patients were followed for progression and survival through December 2011. PFS was defined as the time between study enrollment and progression of disease or death. In the analysis of PFS, patients who withdrew from the study for reasons other than progression or death were censored at the time of discontinuation of study therapy. OS was defined as the time between study enrollment and death due to any cause. Both PFS and OS were estimated by the Kaplan-Meier method.
Biochemical response was defined as a decrease in chromogranin A by 50% or more from baseline in patients with an elevated chromogranin A at baseline. Toxicity and complications of treatment were assessed based on reports of adverse events, physical examinations, and laboratory assessments.
This study used 2 cohorts. The first cohort received temozolomide, 150 mg/m2 on days 1 to 7 and days 15 to 21 of a 28-day cycle, in combination with everolimus at a dose of 5 mg daily, and the second cohort received temozolomide in combination with everolimus at 10 mg daily if treatment was tolerated in cohort 1. Power calculations were based on the anticipation that patients were receiving full doses of both drugs in cohort 2. The proposed regimen was to be considered worthy for further investigation if a true response rate in this cohort was 15% or greater and not worthy if it was 5% or less. A total of 32 evaluable patients were planned in cohort 2. Seventeen patients were entered in the first stage; one response was required to enroll an additional 15 patients. With this design, the study had overall power of 83% and an overall type 1 error of 0.18.
A total of 43 patients enrolled in the study, including 7 in cohort 1 and 36 in cohort 2 (Table 1). Patients had a median age of 53 years; 60% were male, and approximately 40% received concurrent treatment with octreotide. The majority (77%) had not received prior systemic therapy, with the exception of octreotide. Thirty patients (70%) had evidence of radiographic progression of disease prior to initiation of therapy.
Table 1. Baseline Characteristics of the Patient Population
Other regimens include sorafenib (n = 1), pazopanib (n = 2), endostatin (n = 1), and interferon (n = 1).
No. of subjects
Cohort 1: 7 Cohort 2: 36
Median age, y (range)
26 (60)/17 (40)
ECOG performance status (0;1)
20 (47); 23 (53)
Median baseline chromogranin A, ng/mL (range)
Patients receiving concurrent octreotide
Patients receiving prior octreotide
Patients with progressive disease prior to study entry
Number of prior systemic therapy regimens (other than octreotide)
Seven patients, of whom 6 were evaluable, were enrolled in cohort 1 and received temozolomide at a dose of 150 mg/m2 daily for 7 days every other week in combination with everolimus at 5 mg daily. Among the first 3 patients, 1 patient experienced a DLT consisting of grade 4 thrombocytopenia. Four additional patients, of whom 3 were evaluable, enrolled with no further observed DLT, and cohort 2 was subsequently initiated. Thirty-six patients enrolled in cohort 2 and were treated with temozolomide at 150 mg/m2 daily for 7 days every other week in combination with everolimus at 10 mg daily.
Patients in the 2 cohorts received a median of 8.5 four-week treatment cycles (range, 1-28). Twenty-one patients discontinued treatment due to disease progression, which was the most common reason for treatment discontinuation (Table 2). An additional 9 patients discontinued therapy due to treatment-related toxicity, including pneumonitis (n = 3), edema (n = 1), rash (n = 2), elevated creatinine (n = 2), and elevated liver function tests (n = 1). Eight patients were removed at investigator discretion or for withdrawal of consent. At the time of data analysis, 5 patients were still receiving study therapy.
Table 2. Reasons for Treatment Discontinuation
Reason for Treatment Discontinuation
No. (%) n = 43
Disease progression (per RECIST)
Investigator discretion/withdrawal of consent
Elevated liver function tests
Remain on study
Although hematologic toxicities did not lead to any cases of treatment discontinuation, they were nevertheless commonly observed (Table 3). Nearly half (44%) of patients experienced grade 3 or 4 lymphopenia, and 16% experienced grade 3 or 4 thrombocytopenia. However, complications from hematologic toxicities were uncommon. Infections were relatively rare. One patient without prior knowledge of hepatitis B infection experienced reactivation of viral hepatitis (consisting of grade 4 elevation in alanine aminotransferase) after completing 9 months of therapy and while receiving everolimus alone. After discontinuation of everolimus and initiation of therapy for hepatitis B, his liver function tests normalized. Pneumonitis developed in 3 (7%) patients. There were no cases of bleeding or hemorrhage. Other common toxicities included mild mucositis, hyperglycemia, hypercholesterolemia, or hypertriglyceridemia. Electrolyte disturbances, including decreased magnesium, calcium, and phosphate levels, were also observed but did not cause symptoms.
Among 40 patients who underwent restaging scans and were evaluable for response, 16 (40%) experienced partial responses, as defined by RECIST criteria. Nearly all of the remaining patients (53%) experienced stable disease as their best response to therapy, and only 3 patients (7%) had progressive disease. Biochemical responses, defined as a > 50% decline in chromogranin A from baseline in patients with elevated baseline levels, were also common and were observed in 14 of 31 (45%) patients with elevated baseline levels (Table 4).
Table 4. Efficacy of Everolimus and Temozolomide
A total of 31 patients had elevated baseline chromogranin A levels above the upper limit of the reference range of ≤225 ng/mL and were evaluable for biochemical response.
No. of patients with baseline elevated chromogranin A level who experienced a >50% decrease
No. of patients with baseline elevated chromogranin A without subsequent biochemical assessment
The median PFS duration was 15.4 months (95% confidence interval = 9.4-20.4 months) months; median OS was not reached (Fig. 1).
Previous studies have demonstrated that temozolomide and everolimus independently have activity in patients with neuroendocrine tumors. In this study, we demonstrate that these 2 agents can be administered together safely at therapeutic dose levels. We further show that this regimen is active and associated with both a high response rate and with encouraging PFS duration.
Temozolomide-based regimens previously have been shown to be active in patients with advanced pancreatic NET. Estimated overall response rates in retrospective studies range from 8% to 70%.[8-11] Prospective studies have suggested overall tumor response rates of 33% to 60%.[12-14] These studies are limited, however, by the inclusion of small numbers of patients with pancreatic NET with response estimates based on 15 patients or fewer. Our observed response rate of 40% in 40 patients with pancreatic NET is consistent with previous observations and serves to confirm the activity of temozolomide-based regimens in this setting.
Two biologically targeted agents, sunitinib and everolimus, were recently approved for use in patients with advanced pancreatic NET based on the results of large, placebo-controlled studies.[1, 2] Overall tumor response rates in these studies as measured by RECIST, however, were low. Sunitinib was associated with an overall response rate of 9%, and everolimus with an overall response rate of 4%. The primary benefit observed with these agents was their effect on PFS. Treatment with sunitinib was associated with a median PFS of 11 months, as compared with 4.5 months for placebo; treatment with everolimus was associated with a median PFS of 11.4 months, as compared with 5.5 months for placebo.
The median PFS of 15.4 months observed in this study with the combination of temozolomide and everolimus is encouraging and superior to the reported PFS durations observed with everolimus or sunitinib in the randomized studies. These results, however, need to be interpreted with caution given the fact that this is a single-arm study. In addition, a requirement for disease progression prior to study enrollment was included in the randomized registration studies of everolimus and sunitinib, but was not included in our study. Although the majority of patients in our study (70%) had evidence of disease progression prior to therapy, some patients, particularly those with symptomatic or extensive disease, began therapy shortly after diagnosis. The inclusion of patients without prior evidence of progression may overestimate PFS due to the often indolent nature of disease.
There also were differences among patients in use of concurrent octreotide that could potentially influence interpretation of our study results. However, given that objective tumor shrinkage with octreotide is rare, it is unlikely that octreotide use has influenced our observations regarding the response rate associated with temozolomide and everolimus. Furthermore, although octreotide has been associated with improved PFS in midgut NETs, an antiproliferative effect in pancreatic NET has not been established to date.
A potential concern in administering temozolomide and everolimus together is the possibility of cumulative toxicity. To reduce this risk, we limited the duration of temozolomide to 6 months; in addition, all patients received prophylaxis for pneumocystis pneumonia, given prior reports of lymphopenia associated with temozolomide as well as the potential immunosuppressive effect of everolimus.[1, 16] Encouragingly, we saw no serious opportunistic infections with this regimen. We did observe one case of hepatitis B reactivation in a patient not previously known to have hepatitis B infection. Reactivation of hepatitis B has been associated with the immunosuppressive effects of everolimus, and screening should be considered for patients with risk factors for infection with hepatitis B.
Although most patients tolerated this regimen without difficulty, we did observe many of the anticipated toxicities of each drug. These included lymphopenia and thrombocytopenia presumably related primarily to temozolomide, as well as anticipated side effects of everolimus such as hyperglycemia, mild mucositis, and, less commonly, pneumonitis. It did not appear that the incidence of these adverse events was any higher than would be expected with either drug alone.
The optimal dosing schedule of temozolomide for the treatment of advanced pancreatic NET remains an open question. Our study represents the third prospective study using a dose-intense regimen of temozolomide administered at a dose of 150 mg/m2 daily for 7 days every other week.[12, 13] Given the risk of long-term hematologic toxicity with this regimen, we limited treatment duration with temozolomide to 6 months. A commonly used combination of temozolomide and capecitabine has also been shown to be active and uses a schedule in which temozolomide is administered at a dose of 200 mg/m2 for 5 days every 4 weeks. Most recently, activity was also observed with a regimen incorporating low-dose, metronomic temozolomide. Future studies more formally evaluating the optimal dosing regimen for temozolomide in pancreatic NET are warranted. The tolerability of everolimus in combination with dose-intense temozolomide observed in our study, however, suggests that everolimus can also be given in concert with temozolomide using these other dosing regimens.
In conclusion, the randomized studies of sunitinib and everolimus clearly demonstrate benefit associated with these targeted agents in patients with advanced pancreatic NET. Similar randomized data does not exist currently for temozolomide. However, patients with a high tumor burden or symptoms related to bulk of disease may benefit from treatment with a cytotoxic agent due to the higher response rates associated with temozolomide-based regimens and other alkylating agents compared with either sunitinib or everolimus. Our study demonstrates that these patients can be safely and effectively treated with a combination of temozolomide and everolimus. Future studies evaluating the optimal dosing schedule of temozolomide and how much might be gained with regard to response rate and progression-free survival duration with combination therapy compared to treatment with either agent alone are warranted.
This work was supported by Novartis and Schering-Plough/Merck.
CONFLICT OF INTEREST DISCLOSURE
Dr. Chan has received research funding from Novartis, Schering-Plough/Merck, Bayer, and Onyx Pharmaceuticals; and has stock ownership in Merck.