Abstract
Background/Aim: Pancreatic neuroendocrine tumors (panNETs) are rare neoplasms with challenging disease management. We aimed to evaluate the progression-free survival (PFS) and overall response rate (ORR) in chemotherapy-naïve patients with unresectable or metastatic Grade (G) 1-2 panNETs treated with everolimus in the routine care in Greece. Patients and Methods: This was a multicenter, prospective, observational study. Eligible patients were recently (≤4 weeks) initiated on treatment with everolimus and were followed for up to 48 months. Results: Nineteen eligible patients (mean age 55.1 years) were enrolled. All patients had metastatic disease and 84.2% had G2 panNET. Everolimus was initiated in combination with somatostatin analogues in 84.2% of the patients. The mean everolimus treatment duration was 21.5 months. The median Kaplan-Meier-estimated PFS was 20.4 months (95% confidence interval=14.1-41.5). The ORR was 27.8%. The rate of everolimus-related adverse events was 84.2% (Grade ≥3: 31.6%). Conclusion: Everolimus displayed clinical benefit and a predictable safety profile in pancreatic neuroendocrine tumors.
- Effectiveness
- everolimus
- pancreatic neuroendocrine tumors
- progression-free survival
- real-world study
- tolerability
Pancreatic neuroendocrine tumors (panNETs) comprise 1-2% of all pancreatic tumors and less than 10% of all neuroendocrine tumors (1). While panNETs are considered rare, a steady increase in their annual incidence has been observed over the last decades (2–4). In particular, according to the Surveillance, Epidemiology, and End Results program the annual incidence of panNETs increased from 0.4 per 100,000 in 2004 to 1.0 per 100,000 in 2016 (2), while according to a population-based analysis in Canada it increased from 0.3 per 100,000 in 2004 to 0.6 per 100,000 in 2009 (2, 3, 5). The increased incidence seems to be mainly attributable to early diagnoses of patients with localized disease and grade 1 (G1) tumors (2, 5).
PanNETs arise from the endocrine tissue of the pancreas; yet, the specific cell type of origin may vary, which accounts for the large heterogeneity of these tumors with respect to their histological and secretory properties, and, consequently, their clinical presentation. The symptoms of panNETs can be due to hormone over-secretion, including among others insulin (insulinoma), gastrin (gastrinoma), vasoactive intestinal peptide (VIPoma), and glucagon (glucagonoma). On the other hand, between 60%-90% of all panNETs do not secrete any known hormone and are characterized as non-functional (6). These tumors are often discovered due to compressive symptoms to nearby structures and/or the presence of distant metastases (7).
PanNETs are generally classified by the degree of differentiation and grade (G), which is based on the histological assessment estimation of the mitotic count and the Ki67 proliferative index. G1 and G2 panNETs are well differentiated and exhibit a relatively slow growth depending on the Ki67 value. G3 panNETs have higher proliferative capacity and can be either well-differentiated and are designated as G3-NET, or poorly differentiated, in which case they are designated as neuroendocrine carcinomas (NEC) and tend to be more aggressive (3–10). The main treatment approach for panNETs is surgical resection. However, in patients with metastatic and/or advanced disease, as well as in patients with comorbidities that do not allow surgical interventions, a multimodal approach utilizing medical, radionuclear and/or cytoreductive techniques is employed (11–13). Debulking surgery is considered beneficial for patients with metastatic functional panNETs of high tumour burden, and difficult to control relevant symptoms (14). Several systemic therapy options are currently available, including long acting somatostatin analogues (SSAs) (e.g., octreotide, lanreotide), radiolabeled SSAs, antiangiogenic agents (e.g., bevacizumab), interferon-α, chemotherapy, and targeted therapies (everolimus, sunitinib, surufatinib) (11, 14). However, the optimal treatment algorithm along with the ideal sequence of treatments for advanced/metastatic disease is not precisely defined although recommendations in specific clinical settings have recently been developed (14). Adjuvant therapy with platinum-based chemotherapy can be considered only for aggressive NEC G3 (14). SSAs are considered standard first-line therapy in patients with functional panNETs, and can be considered as first-line therapy for control of tumour growth in panNETs with Ki67 up to 10% that are somatostatin receptor-positive. Everolimus and sunitinib are both approved for advanced progressive panNETs. In patients with progressive G1/G2 disease and non-resectable liver and/or other distant metastases chemotherapy with streptozotocin and 5-fluorouracil is recommended, while temozolomide alone or in combination with capecitabine represent an alternative chemotherapeutic approach. For patients with NEC G3, the recommended first-line chemotherapy is cisplatin or carboplatin with etoposide. Various chemotherapeutic options are available in the second-line treatment setting (14). Advances in available treatment options have resulted in near doubling of the overall survival (OS) in patients with panNETs and distant metastases in the period of 2009-2012 compared to that of 2000-2004 [hazard ratio=0.56; 95% confidence interval (CI)=0.44-0.70] (2). Everolimus, a mammalian target of rapamycin inhibitor, approved for the treatment of unresectable or metastatic G1 or G2 panNETs in adults with progressive disease, demonstrated a significantly longer progression-free survival (PFS) than placebo in the phase III RADIANT 3 trial (15). Owing to the limited availability of real-world data regarding the effectiveness of everolimus at the time of designing the study, the PROTOR observational study aimed at evaluating the long-term effectiveness, safety, and tolerability profile of everolimus in chemotherapy naïve patients with unresectable or metastatic G1/G2 panNET who were not eligible to receive chemotherapy, and who were routinely managed in Greece.
Patients and Methods
Study design and setting. This was a multicenter, prospective, observational study. Eligible patients were enrolled from six public/university hospital sites distributed across three geographic regions of Greece. The study was designed and conducted in accordance with all applicable local laws and regulations, the Guidelines for Good Pharmacoepidemiology Practices of the International Society for Pharmacoepidemiology, and the ethical principles laid down in the Declaration of Helsinki, and was approved by the Institutional Review Boards of all participating hospital sites (Laiko Hospital, Theagenio Cancer Hospital, Hygeia Hospital, University Hospital of Heraklion, Evangelismos Hospital, and Mitera Hospital). Written informed consent was obtained from all patients.
To control for patient selection bias, consecutive enrollment of patients that met the study specific eligibility criteria and attended the study sites was utilized. The planned recruitment period was 36 months. The follow-up period for each patient started from study enrolment and up to 12 months after the inclusion of the last patient in the study. Premature discontinuation of study participation took place in case of everolimus permanent discontinuation, disease progression, withdrawal of consent, protocol deviation, or death. After the end of the active follow-up period, a 12-month survival follow-up period via two telephone contacts applied to each patient.
Patients. Eligible patients were adults with unresectable or metastatic, G1 or G2 panNET, with progressive disease, who were chemotherapy-naïve. Patients had initiated treatment with everolimus according to the locally approved Summary of Product Characteristics (SmPC) and the standard medical practice within the past four weeks prior to their enrollment into the study. The investigators’ decision to administer everolimus to a patient preceded the consideration of patient’s eligibility for enrollment into the study.
Study objectives and relevant definitions. The primary objective of the study was to evaluate the PFS. The secondary objectives of the study included the evaluation of OS and the overall response rate (ORR) in the study population, as well as the assessment of the safety and tolerability profile of everolimus.
Baseline assessments were defined as those performed at the earliest timepoint before or at everolimus treatment onset. Progression was defined as the time elapsed between everolimus treatment initiation and the date of documented disease progression as per the Response Evaluation Criteria In Solid Tumors (RECIST) v1.1 or date of death from any cause, whichever occurred first. ORR was defined as the proportion of patients whose best response was either complete response (CR) or partial response (PR) as per the RECIST v1.1 criteria.
Data collection. The study involved primary data collection, obtained prospectively during the study visits as performed per standard clinical practice. Information regarding the underlying disease, medical history and prior treatment for panNET was recorded from the patients’ medical records. Data were collected using paper case report forms. Laboratory, imaging tests, and examinations were performed as per the physicians’ standard medical practice. Follow-up visit frequency was determined by the treating physician based on his/her standard practice; however, study-related data were planned to be collected at 3-month intervals during the first year of follow-up and every 6 months thereafter, until completion of study participation. According to the protocol, data to be collected at the study visits included, evaluation of the Eastern Cooperative Oncology Group (ECOG) performance status, laboratory parameters, and imaging (computed tomography or magnetic resonance imaging) response assessments, according to the routine procedures of the study site. No central review of disease response assessment was performed, instead response assessment was performed locally by the investigators.
Sample size estimation. The planned sample size was based on practical considerations and feasibility projections without formal calculation. According to an annual incidence of panNET of 3.5 to 4 cases per million (16) and the 11 million-population of Greece, approximately 38 to 44 new diagnoses were estimated to occur every year in the country. Assuming that 30-45% of those patients would be eligible for participation, a study size of 40 patients was planned to be enrolled over the 36-month recruitment period.
Statistical analysis. Statistical analysis was performed using SAS® software v.9.4 (SAS Institute Inc., Cary, NC). Summary statistics of continuous variables are presented as mean [standard deviation (SD)], or as median [interquartile range (IQR), or range]. Regarding the primary endpoint analysis, the Kaplan-Meier median (95% CI) time to disease progression/death was estimated. Patients with no documented disease progression/death during the observation period were censored at the date of their last assessment, denoting absence of progression. Regarding estimation of OS, the planned time-to-event analysis was not performed as only one death was observed during the course of the study. For assessment of the ORR, the number (%) of patients with CR and/or PR, along with the 95% Clopper-Pearson CI was calculated. No imputation methods were performed in regards to missing data, with the exception of partial dates.
Results
Patient disposition. A total of 19 eligible patients were enrolled in the study between September 2013 and September 2016. Overall, 78.9% (15/19) of the patients prematurely discontinued study participation mainly due to disease progression (66.7%; 10/15) (Figure 1). The median length of patient follow-up was 20.0 months (range=3.2-47.9 months), while the median time to premature discontinuation of study participation was 17.0 months (range=3.2-41.6 months).
Flow-chart of patient enrollment, eligibility, follow-up and withdrawal from the study.
Patient baseline characteristics. All eligible patients were Caucasian and 57.9% (11/19) were male. The patients’ mean (SD) age at everolimus treatment initiation (baseline) was 55.1 (12.5) years. Moreover, at baseline, 89.5% (17/19) of the patients had an ECOG performance status of 0 (Table I).
Baseline demographics and disease characteristics of the overall population.
PanNET characteristics and prior therapies. Patients were initially diagnosed with a panNET at a median (IQR) age of 53.7 (47.8-60.3) years. PanNETs were non-functional in 77.8% (14/18) of patients with available data, and functional in 22.2% (4/18), of whom one patient had an insulinoma and three patients gastrinomas. According to the WHO grading system, 84.2% (16/19) of the patients had G2 and 15.8% (3/19) G1 panNET. At enrollment all patients had metastatic disease. The most common metastatic site was the liver (Table I). In total, 89.5% (17/19) of the patients had received therapy for panNET management prior to everolimus initiation.
Everolimus treatment characteristics and concomitant medications for panNET management. Everolimus was initiated at or within 28 days prior to enrollment (median: 3 days prior to enrollment). All patients were initiated on a 10 mg everolimus dose, 15.8% (3/19) as monotherapy and 84.2% (16/19) as combination therapy with SSAs (Table II). In addition to the patients who were initiated on everolimus in combination with SSAs, one patient started receiving an SSA (octreotide long-acting release [LAR]), during the course of the study, leading to 89.5% (17/19) of the patients in total having received everolimus in combination with SSAs. Moreover, during the course of the study, one patient who was initiated on everolimus in combination with lanreotide switched to a combination with octreotide LAR.
Everolimus treatment characteristics.
The mean (SD) treatment duration was 21.5 (13.7) months, and the mean (SD) exposure (i.e., treatment duration minus temporary interruptions) was 21.0 (13.6) months. Overall, 68.4% (13/19) of the patients underwent dose modifications and/or treatment interruptions throughout the study observation period. In particular, 63.2% (12/19) of the patients underwent a total of 19 interruptions, and 31.6% (6/19) of the patients underwent 10 dose reductions (Table II).
Treatment with everolimus was permanently discontinued in 52.6% (10/19) of the patients after a mean (SD) of 16.2 (8.5) months. The reasons for permanent treatment discontinuation included disease progression/relapse in six patients, adverse event/toxicity (other than progressive disease) in four patients (one of whom subsequently experienced disease progression, and one of whom died without documented disease progression). Everolimus discontinuation was not recorded in three patients who experienced disease progression, as data collection ceased and the patients were withdrawn from the study (as per protocol). For one of these patients it became known (after withdrawal from the study) that everolimus treatment would be continued until initiation of the next line of therapy.
Progression-free survival and overall response rate. During the course of the study 10 patients experienced disease progression and one patient died without documented disease progression. The adverse events with fatal outcome experienced by this patient included blood glucose increase assessed as being related to everolimus, and fever, hypercalcemia, ascites, renal and hepatic failure which were not assessed as being related to everolimus. Over a Kaplan-Meier-estimated median follow-up period post everolimus treatment onset of 46.1 months, the median Kaplan-Meier-estimated PFS was 20.4 months (95% CI=14.1-41.5). The 2- and 3-year Kaplan-Meier estimated PFS rates were 37.5% and 28.1%, respectively (Figure 2). Throughout the course of study participation, response according to the RECIST v.1.1 criteria was evaluable for 94.7% (18/19) patients. The best response was PR for 27.8% (5/18) of the patients and stable disease for 72.2% (13/18). Accordingly, the ORR among the evaluable patients was 27.8% (5/18; 95% CI=9.7-53.5).
Kaplan-Meier survival curve of progression-free survival (PFS), (n=19).
Safety. A total of 90 adverse events experienced by 84.2% (16/19) patients were assessed as being related to everolimus. Of these, 42 events, experienced by 52.6% (10/19) of the patients, were non-serious and 48 events, experienced by 73.7% (14/19), were serious. Ten events (all serious), experienced by 31.6% (6/19) of the patients were of severity Grade≥3. Of the serious adverse events, those reported with a frequency ≥10.0% were diabetes mellitus [by 36.8% (7/19) of the patients]; glycosylated hemoglobin increase [by 21.1% (4/19)]; stomatitis, and blood glucose increase [each experienced by 15.8% (3/19) of the patients]; aspartate aminotransferase increase, dyslipidemia/hypertriglyceridemia, and gamma-glutamyltransferase increase [each experienced by 10.5% (2/19) of the patients].
Discussion
The PROTOR real-world prospective study systematically captured data regarding the outcomes of 19 chemotherapy-naïve patients with metastatic G1/G2 panNET treated with everolimus in routine care settings. The study results support the effectiveness of everolimus and complement the existing evidence on its safety profile.
In PROTOR, the Kaplan-Meier estimated PFS was 20.4 months, which is longer than that reported in the phase II RADIANT 1 and the phase III RADIANT 3 trials. Specifically, in RADIANT 1, median PFS of 9.7 and 16.7 months were reported for patients receiving everolimus as monotherapy or combination therapy, respectively (17), while in RADIANT 3 the PFS in the everolimus plus best supportive care arm was 11.0 months (15). The PFS observed in RADIANT 3 was comparable between patients who had received prior chemotherapy and chemonaïve patients (11.0 vs. 11.4 months) (18). The PFS in PROTOR is also longer than that of 11 months reported by Panzuto and colleagues in a real-world study (19). Moreover, in a retrospective study conducted in Greece, which included 42 patients with well-differentiated panNETs treated with everolimus, PFS was 12 months when everolimus was administered as second line therapy (20). Furthermore, in a real-world study including 31 patients with advanced well-differentiated panNETs and stage IV disease, who were treated with everolimus followed by sunitinib or with sunitinib followed by everolimus, the median PFS under everolimus treatment was 16.3 months when it was received as first line and 15.5 months when it was received as second line therapy (21). The higher observed PFS in PROTOR could be related to differences in the study design, health care settings, follow-up frequency, response-assessment criteria, lack of central review response assessment, differences in disease characteristics of the enrolled populations, and to a high percentage of patients receiving combination treatment with SSA (15, 17). However, an advantage in PFS from adding SSAs to everolimus is not supported from the results of the COOPERATE-2 trial (22), nor from evidence indicating that addition of SSAs to everolimus did not improve the anti-proliferative and anti-secretory activity of everolimus in primary cultures of human panNETs (23). Therefore, the findings of the present study should be interpreted with caution.
On the other hand, the PFS observed in this is study is somewhat shorter than that observed in a Japanese real-world study (27.5 months) (24), and the OBLIQUE real-world study (25.1 months) conducted in the UK (25). Factors that may account for the observed difference between the PROTOR and OBLIQUE studies, include the higher percentages of patients with metastatic disease (100.0% vs. 72.0%) and G2 panNET (84.2%, vs. 60.0%) at baseline in PROTOR compared to the OBLIQUE study, and the use of the RECIST v1.1. criteria versus investigator’s assessment based on clinical and radiological assessments as per routine practice in OBLIQUE (25).
Regarding the estimation of the OS in PROTOR, one death was observed in the study over approximately 22 months of follow-up and the median OS was not reached. This is not surprising as according to recent data the median OS of patients with panNET has reached 3.6 years (2), with a large Canadian population study reporting a 3-year survival rate of 68% (3). Moreover, updated results of RADIANT 3, reported a median OS of 44 months compared to the historical 27 months (26).
The ORR rate in PROTOR was 27.8% (all PRs), higher than that reported in RADIANT 1 (PR rate: 9.6% with everolimus monotherapy and 4.4% with everolimus + octreotide LAR) (17), in RADIANT 3, where the ORR (all PRs) was 5% (15), and in a Japanese real-world study where it was 15.0% (24). In the retrospective study of Daskalakis et al. the disease control rate (i.e., PRs plus stable disease) was 85.7% among patients receiving everolimus as first line and 70% among those receiving everolimus as second line treatment (20). The respective disease control rate in our study was 100% among patients evaluable for a response.
Regarding the adverse events observed in the context of this study, no new safety signals emerged, with most adverse events listed as very common or common in the product’s SmPC. Of the 90 adverse reactions recorded only 10 were Grade≥3, which in combination with the long treatment duration (median: 21.5 months) and the relatively low rate of treatment discontinuation due to adverse events [four patients (21%)], underscores the benefit of treatment modifications aiming to maximize time-on-treatment and the resulting clinical benefit of everolimus. With respect to treatment changes, 63.2%, and 31.6% of patients in PROTOR underwent treatment interruption(s) and dose reduction(s), with the respective percentages being 60.4% and 83.3% in the OBLIQUE study (25). In addition, over a relatively similar treatment duration period between the two studies (median: 21.5 months in PROTOR and 27.8 months in OBLIQUE) a similar percentage of patients permanently discontinued treatment in PROTOR and in OBLIQUE (52.6% and 47.9%, respectively) (25).
Main limitations of the study include its small study size, and follow-up period along with the absence of central radiological review, which restricts the interpretation of the study outcomes. The lower than expected number of patients enrolled in this study may be a result of overestimation of the percent of panNET cases that would be eligible for participation when estimating the sample size. As mentioned above, response assessment is also limited by lack of a central response assessment team, with response lying solely upon local hospital review. Moreover, in accordance with the study’s non-interventional design, the frequency of assessments was not predefined by the protocol and was dependent on the physician’s routine practice. Nonetheless, this study is aimed to generate real-world data in patients with panNET. Overall, the study results support the use of everolimus in the management of advanced panNETs and are in alignment with the accumulating real-world evidence on the effectiveness of everolimus in advanced panNET, even in patients with advanced ECOG stages and in later treatment lines, which is complemented by a manageable and predictable safety profile and the ability to preserve the patient’s quality of life (20, 21, 25).
In conclusion, the present prospective observational study provides evidence of the effectiveness, safety and tolerability profile of everolimus in metastatic G1/G2 panNET patients who are chemotherapy-naïve. The results demonstrate the beneficial effects of everolimus on PFS prolongation and treatment response accompanied by a predictable and manageable safety and tolerability profile.
Acknowledgements
The PROTOR study was sponsored by Novartis. The study sponsor was involved in the study concept and design.The authors would like to acknowledge Qualitis Ltd. CRO, Greece for medical writing support, funded by Novartis Hellas.
Footnotes
Authors’ Contributions
All Authors contributed to the implementation of the study, reviewed and provided their input on the manuscript writing.
Conflicts of Interest
All Authors declare that there are no relevant competing interests to report.
- Received December 6, 2021.
- Revision received February 16, 2022.
- Accepted February 18, 2022.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.