Guideline A Quality Initiative of the Program in Evidence-Based Care (PEBC), Cancer Care Ontario (CCO)

Transcription

1 A Quality Initiative of the Program in Evidence-Based Care (PEBC), Cancer Care Ontario (CCO) Systemic Therapy of Incurable Gastroenteropancreatic Neuroendocrine Tumours S. Singh, D. Sivajohanathan, T. Asmis, C. Cho, N. Hammad, C. Law, R. Wong, K. Zbuk and the Gastrointestinal Disease Site Group Report Date: December 9, 2016 For information about this document, please contact Dr. Simron Singh, the lead author, through the PEBC via: Phone: ext Fax: For information about the PEBC and the most current version of all reports, please visit the CCO website at or contact the PEBC office at: Phone: ext Fax: PEBC Report Citation (Vancouver Style): S. Singh, D. Sivajohanathan, T. Asmis, C. Cho, N. Hammad, C. Law, R. Wong, K. Zbuk. Systemic therapy of incurable gastroenteropancreatic neuroendocrine tumours. Toronto (ON): Cancer Care Ontario; 2016 December 9. Program in Evidence-Based Care Guideline No.: 2-21.

2 PUBLICATIONS RELATED TO THIS REPORT Copyright This report is copyrighted by Cancer Care Ontario; the report and the illustrations herein may not be reproduced without the express written permission of Cancer Care Ontario. Cancer Care Ontario reserves the right at any time, and at its sole discretion, to change or revoke this authorization. Disclaimer Care has been taken in the preparation of the information contained in this report. Nevertheless, any person seeking to consult the report or apply its recommendations is expected to use independent medical judgment in the context of individual clinical circumstances or to seek out the supervision of a qualified clinician. Cancer Care Ontario makes no representations or guarantees of any kind whatsoever regarding the report content or its use or application and disclaims any responsibility for its use or application in any way.

3 Table of Contents Section 1: Recommendations... 1 Section 2: Guideline Recommendations and Key Evidence... 3 Section 3: Guideline Methods Overview... 7 Section 4: Systematic Review Section 5: Internal and External Review References Appendix 1: Affiliations and Conflict of Interest Declarations Appendix 2: Literature Search Strategy Appendix 3: PRISMA Flow Diagram Appendix 4: Quality Assessment of Randomized Controlled Trials Appendix 5: Trials Including and Reporting on Patients with pnets and Non-pNETs... 67

4 Systemic Therapy of Incurable Gastroenteropancreatic Neuroendocrine Tumours Section 1: Recommendations This section is a quick reference guide and provides the guideline recommendations only. For key evidence associated with each recommendation, see Section 2. GUIDELINE OBJECTIVES To make recommendations with respect to systemic therapy for the treatment of patients with incurable gastroenteropancreatic (GEP) neuroendocrine tumours (NETs). TARGET POPULATION Adults with a diagnosis of incurable GEP NETs. INTENDED USERS Clinicians involved in the treatment of patients with GEP NETs. Note: This guideline does not address peptide receptor radionuclide therapy. This is covered in Guideline Further, this guideline addresses anti-proliferative treatment and not symptomatic treatment. RECOMMENDATIONS, KEY EVIDENCE, AND INTERPRETATION OF EVIDENCE Recommendation 1 Patients with well or moderately differentiated pancreatic neuroendocrine tumours (pnets) should be offered targeted therapy (i.e., everolimus or sunitinib). No evidence-based recommendation can be made for or against other types of targeted therapy, somatostatin analogues, chemotherapy or combination therapy due to insufficient evidence. Qualifying Statements for Recommendation 1 Based on previously established trials, standard of care, and considerable years of clinical experience, chemotherapy may remain an option although there have been no recent randomized controlled trials conducted to validate this. Methodologically strong evidence for chemotherapy does not exist, unlike other systemic therapy options. The various systemic therapy regimens, doses, and schedules have not been directly compared with one another, resulting in insufficient data to recommend one over the other. However, targeted therapy (i.e., everolimus or sunitinib) is associated with the largest benefit (i.e., lowest hazard ratio [HR]) for systemic therapy. There is no evidence to support the use of dual biological therapy. Subgroup analysis has shown a strong trend toward benefit in the use of somatostatin analogues in pnets although the HR was not statistically significant. Over-interpretation of these results is cautioned as the subgroup analysis was not adequately powered and contained a low number of events. However, the overall study population did show a significant benefit for SSAs and was comprised of approximately 45% pnet patients [1]. Section 1: Recommendations December 9, 2016 Page 1

5 Recommendation 2 Patients with non-pnets should be offered either targeted therapy (i.e., everolimus) or somatostatin analogues (i.e., octreotide long-acting repeatable [LAR] or lanreotide). No evidence-based recommendation can be made for or against other types of targeted therapy, somatostatin analogs, chemotherapy, or combination therapy due to insufficient evidence. Qualifying Statements for Recommendation 2 The evidence for everolimus is specific to patients with non-functional tumours and is based on a subgroup analysis of gastrointestinal patients, although the results of a preceding trial suggest some benefit for patients with functional tumours receiving everolimus with octreotide LAR. However, there is uncertainty surrounding these results as this trial did not meet its pre-specified endpoint for analysis and contained a small percentage of patients with pancreatic and lung NETs. The various targeted therapy and somatostatin analogue regimens, doses, and schedules have not been directly compared with one another, resulting in insufficient data to recommend one over the other or any preferred method of sequencing. Section 1: Recommendations December 9, 2016 Page 2

6 Systemic Therapy of Incurable Gastroenteropancreatic Neuroendocrine Tumours Section 2: Guideline Recommendations and Key Evidence GUIDELINE OBJECTIVES To make recommendations with respect to systemic therapy for the treatment of patients with incurable gastroenteropancreatic (GEP) neuroendocrine tumours (NETs). TARGET POPULATION Adults with a diagnosis of incurable GEP NETs. INTENDED USERS Clinicians involved in the treatment of patients with GEP NETs. Note: This guideline does not address peptide receptor radionuclide therapy. This is covered in Guideline Further, this guideline addresses anti-proliferative treatment and not symptomatic treatment. RECOMMENDATIONS, KEY EVIDENCE, AND INTERPRETATION OF EVIDENCE Recommendation 1 Patients with well or moderately differentiated pancreatic neuroendocrine tumours (pnets) should be offered targeted therapy (i.e., everolimus or sunitinib). No evidence-based recommendation can be made for or against other types of targeted therapy, somatostatin analogues, chemotherapy or combination therapy due to insufficient evidence. Qualifying Statements for Recommendation 1 Based on previously established trials, standard of care and considerable years of clinical experience, chemotherapy may remain an option although there have been no recent randomized controlled trials (RCTs) conducted to validate this. Methodologically strong evidence for chemotherapy does not exist, unlike other systemic therapy options. The various systemic therapy regimens, doses, and schedules have not been directly compared with one another, resulting in insufficient data to recommend one over the other. However, targeted therapy (i.e., everolimus or sunitinib) is associated with the largest benefit (i.e., lowest hazard ratio [HR]) for systemic therapy. There is no evidence to support the use of dual biological therapy. Subgroup analysis has shown a strong trend toward benefit in the use of somatostatin analogues (SSAs) in pnets although the HR was not statistically significant. Overinterpretation of these results is cautioned as the subgroup analysis was not adequately powered and contained a low number of events. However, the overall study population did show a significant benefit for SSAs and was comprised of approximately 45% pnet patients [1]. Key Evidence for Recommendation 1 The overall quality of the evidence was assessed using the GRADE criteria. The best evidence for targeted therapy comes from two RCTs [2,3]. Based on these two trials, the overall quality of the evidence was deemed moderate, marked down for risk of bias due to patients being able to cross over to the treatment arm at disease progression in both trials. The overall quality of the evidence for SSAs is low due to imprecision (wide confidence Section 2: Guideline December 9, 2016 Page 3

7 interval [CI]) and risk of bias (subgroup not powered for analysis); for chemotherapy is very low due to risk of bias (prospective and retrospective single-arm studies) and imprecision (low patient numbers); and for combination therapy is moderate due to imprecision (wide CI). The best evidence for combination therapy comes from one RCT, in abstract form [4], which was used in assessing the quality. The RADIANT-3 trial, which evaluated the use of everolimus, reported a statistically significant benefit for progression-free survival (PFS) when compared with placebo (HR, 0.34; 95% CI, 0.26 to 0.44; p<0.001) [3]. In addition to the RADIANT-3 trial, median PFS for patients with pnets receiving everolimus ranged from 7.6 months 5.52 to 7.62) to 14 months [3,5-9]. There was no difference in overall survival (OS) between the arms of the RADIANT-3 trial; however, patients receiving placebo were allowed to cross over to the treatment arm after disease progression, confounding the results. A rank-preserving structural failure analysis was later performed to correct for crossover bias (HR, 3.27; 95% CI, 0.10 to13.93) [10]. The p-value was not reported. A phase III randomized trial, which evaluated the use of sunitinib, also reported a statistically significant benefit for PFS when compared with placebo (HR, 0.315; 95% CI, to 0.546; p<0.01) [2]. After a five-year follow-up, a statistically significant OS benefit was shown (HR, 0.40; 95% CI, 0.23 to 0.71; p=0.001) when patients who crossed over to the treatment arm after disease progression were censored [11]. In a matching-adjusted indirect comparison of patients from the RADIANT-3 trial and the phase III sunitinib trial, no statistically significant difference in PFS (p=0.578) and OS (p=0.383) was found for everolimus compared with sunitinib [12]. Preliminary findings of both randomized phase II CALGB80701 and COOPERATE-2 trials, which are only available in abstract form, show no statistically significant differences in PFS between patients receiving everolimus plus bevacizumab compared with everolimus (HR, 0.80; 95% CI, 0.55 to 1.17; p=0.12) [6], and between patients receiving everolimus plus pasireotide compared with everolimus (HR, 0.99; 95% CI, 0.64 to 1.54) [4]. A subgroup analysis of patients with pnets in the CLARINET trial reported no benefit for lanreotide with respect to PFS (HR, 0.58; 95% CI, 0.32 to 1.04) [13]. The p-value was not reported. The overall study, however, showed a significant PFS benefit with treatment (HR, 0.47; 95% CI, 0.30 to 0.73; p<0.001) and included approximately 45% pnet patients[1]. Interpretation of Evidence for Recommendation 1 There was agreement among the members of the Working Group that the overall certainty of the evidence was high for targeted therapy but low for chemotherapy, SSAs, and combination therapy. The Working Group felt that evidence with low overall certainty was insufficient to make definitive recommendations. Although the Working Group looked at PFS, OS, objective response rate (ORR), adverse events, and quality of life, PFS was considered to be the most important outcome, followed by OS and adverse events. Some patient input was sought and quality of life was identified as the most important outcome to patients. However, not all trials routinely collected and reported on quality of life data. In the trials that did report these data, no significant differences were found between the treatment and control arms or between pre- and post-treatment. As such, the Working Group agreed that patients would value increased PFS benefit. The desirable effect (i.e., increased PFS) is considerable, while the undesirable effects (i.e., adverse events) are acceptable for targeted therapy. As a result, the desirable effect is large relative to the undesirable effects. This evidence is generalizable to all patients with pnets. Section 2: Guideline December 9, 2016 Page 4

8 Recommendation 2 Patients with non-pnets should be offered either targeted therapy (i.e., everolimus) or somatostatin analogues (i.e., octreotide long-acting repeatable [LAR] or lanreotide). No evidence-based recommendation can be made for or against other types of targeted therapy, somatostatin analogs, chemotherapy, or combination therapy due to insufficient evidence. Qualifying Statements for Recommendation 2 The evidence for everolimus is specific to patients with non-functional tumours and is based on a subgroup analysis of gastrointestinal (GI) patients, although the results of a preceding trial suggest some benefit for patients with functional tumours receiving everolimus with octreotide LAR. However, there is uncertainty surrounding these results as this trial did not meet its pre-specified endpoint for analysis and contained a small percentage of patients with pancreatic and lung NETs. The various targeted therapy and somatostatin analogue regimens, doses, and schedules have not been directly compared with one another, resulting in insufficient data to recommend one over the other or any preferred method of sequencing. Key Evidence for Recommendation 2 The quality of the evidence was assessed using the GRADE criteria. The best evidence for targeted therapy comes from one RCT [14], which was used to determine the overall quality of the evidence. This RCT was considered to be of moderate quality as it was marked down for risk of bias (subgroup analysis not powered for analysis). The best evidence for SSAs comes from two RCTs [1,15], which were used to determine the overall quality of the evidence. These RCTs were considered to be of low quality as they were marked down for imprecision (low patient numbers) and risk of bias (subgroup not powered for analysis). Although the quality of the evidence may be low, the magnitude of the benefit is large for these trials. The RADIANT-4 trial demonstrated benefit for everolimus in the GI subgroup of patients with non-functional NETs, with a median PFS of 13.1 months 9.2 to 17.3) and 5.4 months 3.6 to 9.3) in the treatment and control arms, respectively (HR, 0.56; 95% CI, 0.37 to 0.84) [14]. The p-value was not reported. Two studies reported on octreotide. In a planned interim analysis for the PROMID study by Rinke et al., a significant benefit in tumour progression or tumour-related death, considered as a surrogate for PFS, was demonstrated for the treatment arm over the control arm (HR, 0.34; 95% CI, 0.20 to 0.59; p= ) [15]. However, no difference in survival was found (HR, 0.83; 95% CI, 0.47 to 1.46; p=0.51) [16]. Meanwhile, a retrospective study reported a median PFS of 51.0 weeks 26.4 to 75.6) for octreotide [17]. The RADIANT-2 trial suggested some PFS benefit for patients with functional tumours receiving everolimus plus octreotide LAR (HR, 0.77; 95% CI, 0.59 to 1.00; p=0.026) [18], although the CI reaches one. Furthermore, this trial did not meet its pre-specified endpoint for analysis and contained a small percentage of patients with pancreatic and lung NETs. A subgroup analysis of patients with midgut NETs in the CLARINET trial reported PFS benefit for lanreotide (HR, 0.35; 95% CI, 0.16 to 0.80; p=0.009) [1,19]. A subgroup analysis of patients with hindgut NETs did not show any benefit for lanreotide, which could be a result of the very low number of patients in this subgroup (n=14). No studies were identified in the literature search that specifically assessed or reported Section 2: Guideline December 9, 2016 Page 5

9 on the role of chemotherapy in patients with non-pnets. Interpretation of Evidence for Recommendation 2 There was agreement among the members of the Working Group that the overall certainty of the available evidence was moderate to high for both targeted therapy and SSAs, but low for combination therapy. Although the Working Group looked at PFS, OS, ORR, adverse events, and quality of life, PFS was considered to be the most important outcome, followed by OS and adverse events. Some patient input was sought and quality of life was identified as the most important outcome to patients. However, not all trials routinely collected and reported on quality of life data. In the trials that did report these data, no significant differences were found between the treatment and control arms or between pre- and posttreatment. As such, the Working Group agreed that patients would value increased PFS benefit. The desirable effect (i.e., increased PFS) is large for both everolimus and octreotide, while the undesirable effects (i.e., adverse events) are small. The Working Group believes the desirable effect is large relative to the undesirable effect. The Working Group believes that although the evidence from the subgroup analyses and planned interim analyses may not be strong, the rarity of this disease coupled with the difficulty of conducting methodologically sound trials in this population result in the need to use the best available evidence to make treatment decisions. This evidence is generalizable to all patients with non-pnets. IMPLEMENTATION CONSIDERATIONS The Working Group considered these recommendations to be the best possible recommendations given the currently available data and their feasibility of implementation. Research has shown that patients in rural settings have poorer outcomes and standardizing care would reduce this inequity. These recommendations would validate and align with what providers are currently implementing. Funding of drugs for NETs must take into account the difficulty in conducting trials with homogeneous populations in this disease and the need to often have heterogeneous populations in order to feasibly assess new systemic therapies. Due to the limited number of cases, funding bodies must recognize that data obtained in the assessment of systemic therapy for NETs are unlikely to be the same level of quality as in other cancers. Accordingly, treatment options that have a biological rationale, such as the use of targeted therapy in the second-line treatment of pnets, should be considered. The Working Group believed the outcomes valued in this guideline would align with patient values and that patients would view these recommendations as acceptable. RELATED GUIDELINES Gulenchyn KY, Yao X, Asa SL, Singh S, Law C. Radionuclide therapy for neuroendocrine malignancies. Toronto (ON): Cancer Care Ontario; 2011 Aug 15. Program in Evidence Based Care Evidence Based Series No.: Section 2: Guideline December 9, 2016 Page 6

10 Systemic Therapy of Incurable Gastroenteropancreatic Neuroendocrine Tumours Section 3: Guideline Methods Overview This section summarizes the methods used to create the guideline. For the systematic review, see Section 4. THE PROGRAM IN EVIDENCE-BASED CARE The Program in Evidence-Based Care (PEBC) is an initiative of the Ontario provincial cancer system, Cancer Care Ontario (CCO). The PEBC mandate is to improve the lives of Ontarians affected by cancer through the development, dissemination, and evaluation of evidence-based products designed to facilitate clinical, planning, and policy decisions about cancer control. The PEBC supports the work of Guideline Development Groups (GDGs) in the development of various PEBC products. The GDGs are composed of clinicians, other healthcare providers and decision makers, methodologists, and community representatives from across the province. The PEBC is a provincial initiative of CCO supported by the Ontario Ministry of Health and Long-Term Care (OMHLTC). All work produced by the PEBC is editorially independent from the OMHLTC. JUSTIFICATION FOR GUIDELINE NETs are the second most prevalent GI cancer after colon cancer due to longer survival periods even in patients with incurable and metastatic disease. Most patients with NETs will undergo various systemic therapies, thus creating the need for an evidence-based guideline. Such guidelines have been lacking and may account for the wide variation in the clinical treatment of NETs. The production of evidence-based guidelines can form the basis of caseby-case discussion by experts in NETs care within a multidisciplinary setting, which is recommended as the optimal treatment pathway in Ontario. GUIDELINE DEVELOPERS This guideline was developed by the Gastrointestinal Disease Site Group (GI DSG) (Appendix 1), which was convened at the request of the Disease Pathway Management Group. The project was led by a small Working Group of the GI DSG, which was responsible for reviewing the evidence base, drafting the guideline recommendations, and responding to comments received during the document review process. The Working Group had expertise in radiation oncology, medical oncology, surgery, and health research methodology. Other members of the GI DSG served as the Expert Panel and were responsible for the review and approval of the draft document produced by the Working Group. Conflict of interest declarations for all GDG members are summarized in Appendix 1, and were managed in accordance with the PEBC Conflict of Interest Policy. GUIDELINE DEVELOPMENT METHODS The PEBC produces evidence-based and evidence-informed guidance documents using the methods of the Practice Guidelines Development Cycle [20,21]. This process includes a systematic review, interpretation of the evidence by the Working Group, draft Section 3: Guideline Methods Overview - December 9, 2016 Page 7

11 recommendations, internal review by content and methodology experts, and external review by Ontario clinicians and other stakeholders. The PEBC uses the AGREE II framework [22] as a methodological strategy for guideline development. AGREE II is a 23-item validated tool that is designed to assess the methodological rigour and transparency of guideline development. The currency of each document is ensured through periodic review and evaluation of the scientific literature and, where appropriate, the addition of newer literature to the original evidence base. This is described in the PEBC Document Assessment and Review Protocol. PEBC guideline recommendations are based on clinical evidence, and not on feasibility of implementation; however, a list of implementation considerations such as costs, human resources, and unique requirements for special or disadvantaged populations is provided along with the recommendations for information purposes. PEBC guideline development methods are described in more detail in the PEBC Handbook and the PEBC Methods Handbook. Search for Existing Guidelines As a first step in developing this guideline, a search for existing guidelines was undertaken to determine whether an existing guideline could be adapted or endorsed. To this end, the following sources were searched for existing guidelines that addressed the research questions: Practice guideline databases: the Standards and Guidelines Evidence Directory of Cancer Guidelines (SAGE), Agency for Healthcare Research and Quality (AHRQ) National Guideline Clearinghouse, and the Canadian Medical Association Infobase. Guideline developer websites: National Institute for Health and Care Excellence (NICE), Scottish Intercollegiate Guidelines Network (SIGN), American Society of Clinical Oncology (ASCO), and the National Health and Medical Research Council - Australia. The following criteria were used to select potentially relevant guidelines: Guideline databases and websites were searched using the following keyword neuroendocrine Only evidence-based guidelines published after 2010 (i.e., less than five years old) were considered. This search did not yield a guideline that could be adapted or endorsed. GUIDELINE REVIEW AND APPROVAL Internal Review For the guideline document to be approved, 75% of the content experts who comprise the GDG Expert Panel must cast a vote indicating whether or not they approve the document, or abstain from voting for a specified reason, and of those that vote, 75% must approve the document. In addition, the PEBC Report Approval Panel (RAP), a three-person panel with methodology expertise, must unanimously approve the document. The Expert Panel and RAP members may specify that approval is conditional, and that changes to the document are required. If substantial changes are subsequently made to the recommendations during external review, then the revised draft must be resubmitted for approval by RAP and the GDG Expert Panel. Section 3: Guideline Methods Overview - December 9, 2016 Page 8

12 External Review Feedback on the approved draft guideline is obtained from content experts and the target users through two processes. Through the Targeted Peer Review, several individuals with content expertise are identified by the GDG and asked to review and provide feedback on the guideline document. Through Professional Consultation, relevant care providers and other potential users of the guideline are contacted and asked to provide feedback on the guideline recommendations through a brief online survey. This consultation is intended to facilitate the dissemination of the final guidance report to Ontario practitioners. Patient- and Caregiver-Specific Consultation Group Four patients/survivors/caregivers participated as Consultation Group members. They reviewed the draft recommendations and provided feedback on its comprehensibility, appropriateness, and feasibility to the Working Group s Health Research Methodologist. The Health Research Methodologist relayed the feedback to the Working Group for consideration. ACKNOWLEDGEMENTS The Working Group would like to thank the following individuals for their assistance in developing this report: Melissa Brouwers, Sheila McNair, Hans Messersmith, Roxanne Cosby, Emily Vella, Eric Winquist, Laurie Elit, and Jonathan Strosberg for providing feedback on draft versions. Terence Tang for conducting a data audit. Sara Miller for copy editing. Section 3: Guideline Methods Overview - December 9, 2016 Page 9

13 Systemic Therapy of Incurable Gastroenteropancreatic Neuroendocrine Tumours Section 4: Systematic Review INTRODUCTION NETs are uncommon malignancies that are located throughout the body. Despite their rarity, NETs are the second most prevalent GI cancer after colon cancer due to longer survival periods even in patients with incurable and metastatic disease. NETs arise from enterochromaffin cells, with the GI tract being the most common primary site, accounting for greater than 60% of NETs [23]. GI NETs can be divided into the clinically relevant entities of pancreatic NETs (pnets) and gastroenterohepatic NETS. NETs are classified as functional when associated with an excessive secretion of hormones or nonfunctional when symptoms derive from the physical manifestations of the tumor. The incidence of NETs is increasing, which can likely be attributed to better detection, improved classification, and increased screening programs that require more widespread use of upper and lower bowel endoscopy, as well as the improved resolution of GI imaging techniques and a heightened awareness of the disease entity. Despite this, the majority of NETs in Ontario present with metastatic disease with no option for cure and with significant impact on quality of life. Recent data from Ontario reported a NET incidence rate of 5.86 cases per 100,000 individuals; a more than a two-fold increase observed over the past 15 years [23]. The prevalence of the disease is also rising due to recent progress in therapeutics as well as earlier identification and, therefore, a longer duration of documented disease. A recent patient experience study documents the considerable burden of disease from NETs, particularly with respect to symptoms, work and daily life, and health care resource use [24]. NETs are an extremely heterogeneous group of tumours with a range of clinical behaviours, making them a difficult group on which to perform high-quality clinical studies. The long natural history of low- to intermediate-grade NETs in particular makes the identification of appropriate trial endpoints challenging. However, in the past 10 years, sufficiently powered therapeutic trials have been successfully conducted so that it is now timely that evidence-based guidelines for management are devised. Guidelines produced by the rigorous evaluation of trials, particularly in a rare tumour type such as NETs, are likely to translate into improved patient care, particularly in geographically large and diverse areas such as Ontario. Furthermore, the measurement of compliance with guidelines can form the basis of practice improvement and professional development. Assimilating new evidence is particularly important for NETs to ensure equity of access to therapies that have been proven to have a significant impact on patient outcomes. The Working Group of the GI DSG developed this evidentiary base to inform recommendations as part of a clinical practice guideline. This review exclusively focuses on the anti-proliferative therapy of GI NETs and does not address the treatment of functional NET symptoms. This review considers the role of the major systemic therapeutic interventions for NETs: SSAs (initially used to control the secretory symptoms of carcinoid syndrome but subsequently shown to have anti-proliferative effect even in non-functional NETs); chemotherapy (which has been poorly studied in low patient numbers with little high-quality data); targeted therapy (including the mtor [mammalian target of rapamycin] inhibitor everolimus and the anti-vegf [vascular endothelial growth factor] agents sunitinib and bevacizumab); and various treatment combinations. This guideline does not address the role of peptide receptor radionuclide therapy, which is covered in Guideline It is noted that Section 4: Systematic Review - December 9, 2016 Page 10

14 most trials did not assess quality of life metrics, which is now recognized as essential in evaluating treatments for incurable cancer. Based on the objectives of this guideline (Section 2), the Working Group derived the research questions outlined below. RESEARCH QUESTION Which of the anti-neoplastic systemic therapies (i.e., chemotherapy, SSAs and interferon α, and targeted therapies [i.e., sunitinib, everolimus, bevacizumab, pazopanib]) is the most effective in improving clinical outcomes (i.e., PFS, OS, ORR, median survival, symptom control, biomarker decreases, quality of life) while minimizing adverse effects (i.e., toxicity) in patients with incurable GEP NETs? METHODS This evidence review was conducted in two planned stages, including a search for systematic reviews followed by a search for primary literature. These stages are described in the subsequent sections. Search for Existing Systematic Reviews A search was conducted for existing systematic reviews. This included original systematic reviews and systematic reviews published as a component of practice guidelines. The MEDLINE (2008 to June 13 th, 2016) and EMBASE (2008 to June 13 th, 2016) databases, as well as the Cochrane Database of Systematic Reviews (2008 to June 13 th, 2016), Cancer Agencies for Drug and Technologies in Health (August 27 th, 2015) and Agency for Healthcare Research and Quality (August 27 th, 2015) were searched for published systematic reviews. The full search strategy is available in Appendix 2. Identified systematic reviews were evaluated based on their clinical content and relevance. Relevant systematic reviews were assessed using the 11-item Assessment of Multiple Systematic Reviews (AMSTAR) [25] tool to determine whether or not it could be incorporated as part of the evidentiary base. Search for Primary Literature In the absence of any relevant systematic reviews, a search was conducted for primary literature. The MEDLINE (2008 to June 13 th, 2016) and EMBASE (2008 to June 13 th, 2016) databases were searched for published phase II and III RCTs and non-rcts. The full search strategy is available in Appendix 2. Reference lists of included primary literature were scanned for additional citations. The following conference proceedings were also searched from 2008 to 2015: ASCO, ASCO Gastrointestinal Cancers Symposium, European Society for Medical Oncology, European Cancer Congress, European Neuroendocrine Tumour Society, and North American Neuroendocrine Tumour Society. Study Selection Criteria and Process Inclusion Criteria 1. Prospective (phase II and III) and retrospective studies with 20 participants. 2. Studies assessing adult patients with incurable GEP NETs. At least 80% of the patients evaluated for the outcomes in each study should have GEP NETs as opposed to other types of NETs (e.g., lung, unknown primary, etc.). 3. Studies that reported on or compared the effects of any of the systematic therapies (i.e., chemotherapy, SSAs or interferon α, and targeted therapies [i.e., sunitinib, everolimus, bevacizumab, pazopanib]) on any of the following clinical outcomes: PFS, Section 4: Systematic Review - December 9, 2016 Page 11

15 OS, ORR, and median survival, with or without biomarker decreases (i.e., chromogranin A, pancreastatin, glucagon), quality of life, and adverse effects. Exclusion Criteria 1. Studies assessing the following conditions: pituitary tumours, large cell neuroendocrine carcinoma, thymic tumours, goblet cell carcinoma, bronchial NETs, paragangliomas, mixed NETs, pheochromocytoma, small cell lung cancer, and thyroid cancer; or 2. Abstracts of non-randomized studies (single-arm clinical trials, case series, etc.); or 3. Abstracts of interim analyses; or 4. Papers or abstracts not available in English; or 5. Letters and editorials that reported clinical trial outcomes; or 6. Papers and abstracts published before A review of the titles and abstracts that resulted from the search was conducted by one reviewer (DS). For items that warranted full-text review, one reviewer (DS) reviewed each item. Data Extraction and Assessment of Study Quality and Potential for Bias Data extraction was conducted by one reviewer (DS) and audited by a second independent auditor (TT). Ratios, including HRs, were expressed with a ratio <1.0 indicating improved efficacy for the experimental arm. Important risks of bias, such as statistical power calculations, sample size, methods of randomization, allocation concealment, blinding, intention to treat analysis, and source of funding were extracted for each randomized study. Criteria from the ROBINS-I tool were used to assess the risk of bias for all non-randomized studies. Criteria from the GRADE method were used to assess the quality of the aggregate evidence for RCTs and non-rcts. Four factors were assessed for each outcome in each comparison: risk of bias, inconsistency, indirectness, and imprecision. Synthesizing the Evidence A meta-analysis was not planned due to the heterogeneity across trials and the inclusion of a large number of single-arm studies. RESULTS Search for Existing Systematic Reviews No relevant systematic reviews were identified. A systematic review on non-surgical treatments for pnets by Valle et al. was found; however, this review was not included due to differences in inclusion criteria (i.e. sample size limits, types of therapy, etc.) [26]. Another systematic review on the role of targeted therapy in metastatic NETs by Lee et al. was found. This review, published in 2016, included only RCTs and performed a meta-analysis of patients treated with targeted therapy [27]. While the review looked at similar patient populations and outcomes as this review, the results were not included for multiple reasons: heterogeneous trials were combined for a meta-analysis, results of patients with pnets and non-pnets were not reported separately, and different drugs under the same class were combined for analysis. Since the purpose of our current systematic review is to inform recommendations, results of RCTs were used as stand-alone evidence and a meta-analysis was not performed due to the heterogeneity across trials. Section 4: Systematic Review - December 9, 2016 Page 12

16 Search for Primary Literature Literature Search Results A total of 10 relevant randomized prospective studies, 16 non-randomized prospective studies, and 12 retrospective studies were found. Table 4-1 summarizes the characteristics of the included studies. A PRISMA flow diagram of the complete search is available in Appendix 3. Where multiple reports and abstracts were published for a single trial or study, only the most recent full publication was included, unless other reports contained data that were not available in the most recent publication. Several trials included both pnets and non-pnets in their populations without presenting results by subgroup. These trials are reported in Appendix 5, but were not used for making recommendations due to heterogeneity and potential for bias. Both pnets and nonpnets respond differently to treatment, thus making it difficult to draw any conclusions from these trials. Although initially determined as outcomes of interest, the Working Group decided not to incorporate biomarker decreases or symptom control in this guideline due to inconsistent and incomplete reporting of this outcome across trials. Section 4: Systematic Review - December 9, 2016 Page 13

17 Table 4-1: Studies Selected for Inclusion Reference Inclusion criteria N Treatment Randomized, prospective trials (11 trials) CALGB80701 Kulke MH (2015) Abstract [6] Phase II Locally advanced or metastatic pnets CLARINET Caplin ME (2014) [1] Well- or moderately differentiated sporadic NETs mg everolimus orally each day VS 10mg everolimus orally each day + 10mg/kg bevacizumab intravenously every 2 wks 120mg lanreotide (Autogel/Depot) subcutaneously every 28 days VS Phan AT (2015) Abstract [13] Phase III RADIANT-2 Pavel ME (2011) [18] [28] Phase III Low-grade or intermediate-grade advanced (unresectable locally advanced or distant metastatic) NETs Placebo every 28 days 10mg oral everolimus once daily + 30mg octreotide LAR every 28 days VS RADIANT-3 Yao JC (2011) [3] Low-grade or intermediate-grade advanced (unresectable or metastatic) pnets Placebo once daily + 30mg octreotide LAR every 28 days 10mg oral everolimus once daily VS Pavel ME (2015) Abstract [10] Phase III RADIANT-4 Singh S (2016) Abstract [29] Yao JC (2016) [14] Phase III Raymond E (2011) [2] Vinik A (2012) Abstract Well-differentiated (G1 or G2), advanced (unresectable or metastatic), non-functional NETs of gastrointestinal or lung origin Well differentiated pancreatic endocrine tumours (advanced, metastatic or both) Placebo once daily *Both arms received best supportive care 10mg oral everolimus once daily VS Placebo once daily Both arms received best supportive care 37.5mg oral sunitinib once daily VS Placebo once daily Section 4: Systematic Review - December 9, 2016 Page 14

18 Reference Inclusion criteria N Treatment [30] *Both arms received best supportive care Raymond E (2016) Abstract [11] Phase III PROMID Rinke A (2009) [15] Rinke A (2016) [16] Phase III SWOG S0518 Yao JC (2015) Abstract [31] Phase III COOPERATE-2 Kulke MH (2015) Abstract [4] Phase II CBEZ235Z2401 Libutti S (2015) Abstract [32] Salazar R (2015) Abstract [8] Phase II Wolin EM (2015) [33] Locally inoperable or metastatic NET and midgut primary tumour or tumour of unknown origin believed to be of midgut origin if a primary within the pancreas, chest, or elsewhere was excluded by multiphasic CT or MRI Metastatic or unresectable, well-differentiated G1 or G2 NETs with progressive disease or other poor prognostic features Advanced, progressive G1 or G2 pnets 81 Advanced, well-differentiated pnets, no prior treatment with an mtor inhibitor and disease progression in the last year Metastatic NET of the digestive systemwho had inadequately controlled carcinoid symptoms mg octreotide LAR intramuscularly every 28 days VS Placebo every 28 days 20mg octreotide LAR + 15mg/kg bevacizumab every 21 days VS 20mg octreotide LAR every 21 days units interferon alpha-2b three times per week 10mg everolimus orally each day VS 10mg everolimus orally each day + 60mg pasireotide intramuscularly every 28 days 400mg BEZ235 twice a day VS 10mg everolimus daily 60mg pasireotide LAR intramuscularly every 28 days VS 57 10mg octreotide LAR intramuscularly every 28 days Non-randomized, prospective studies (16 studies) Ahn HK (2013) Advanced (metastatic) GEP NETs mg oral pazopanib daily with a 28-day cycle [34] Phase II Section 4: Systematic Review - December 9, 2016 Page 15

19 Reference Inclusion criteria N Treatment Advanced NETs mg everolimus once daily Bechter OE (2013) Abstract [5] Pavel M (2013) [35] Phase IIIB Exclusion: Poorly differentiated NET and cytotoxic therapy within 4wks of enrollment BETTER Mitry E (2014) [36] Phase II BETTER Ducreux M (2014) [37] Phase II Brixi-Benmansour H (2011) [38] Phase II Chan JA (2013) [39] Phase I/II Cives M (2015) [40] Phase II Fazio N (2016) [41] Phase II Well-differentiated gastrointestinal tract neuroendocrine carcinoma that had not been previously treated with systemic anticancer therapy Well-differentiated pancreatic neuroendocrine carcinoma that had not been previously treated with systemic anticancer therapy Unresectable, well-differentiated endocrine carcinoma of the pancreas Low- or intermediate grade (G1 or G2) metastatic or locally unresectable pnet Locally unresectable or metastatic G1 or G2 pnets or extrapancreatic NETs Low- or intermediate-grade, unresectable or metastatic pnets with radiological evidence of disease progression since last treatment mg/kg bevacizumab intravenously on day 1 over 30min (± 5min) every 3 wks before chemotherapy mg/m 2 oral capecitabine twice daily at 12hr intervals from day 1-14, followed by a treatment-free period of 7 days mg/kg bevacizumab intravenously on day 1 over 30min every 3 wks before chemotherapy + 400mg/m 2 5-fluorouracil & 500mg/m 2 streptozocin intravenously each day from day 1-5 over 2hrs and repeated every 6 wks 20 FOLFIRI chemotherapy: 180mg/m 2 irinotecan infusion on day 1 with simplified LV5FU2 (a single 2hr infusion of 200mg/m 2 leucovorin on day 1, followed by 400mg/m 2 bolus of 5-fluorouracil, then continuous infusion of 2400mg/m 2 of 5-fluorouracil over 46hrs) 7 Cohort 1: 150mg/m2 temozolomide on days 1-7 and days of a 28-day cycle + 5mg everolimus daily Cohort 2: mg/m2 temozolomide on days 1-7 and days of a 28-day cycle + 10mg everolimus daily 29 60mg pasireotide LAR intramuscularly every 4 wks for 24 months 31 Stage 1: 400mg BEZ235 orally twice daily on a continuous dosing schedule in 28-day cycles Starting dose reduced to 300mg BEZ235 due to tolerability concerns Fjallskog (2008) [42] Phase II ML Stage 2: Best supportive care in combination with either BEZ235 or placebo Never initiated Non-resectable endocrine pancreatic tumours 30 30mg/m 2 liposomal doxorubicin infused intravenously over 1hr on day 1 of each cycle + 1g streptozocin as a bolus injection on days 1-5 during induction course and thereafter, 2g streptozocin on day 1 only Section 4: Systematic Review - December 9, 2016 Page 16

20 Reference Inclusion criteria N Treatment GETNE0801 Castellano D 2013 [43] Phase II Hobday TJ (2015) [44] Phase II Kulke MH (2008) [45] Phase II Kvols LK (2012) [46] Phase II Phan AT (2015) [47] Phase II Yao JC (2010) [9] Phase II Advanced NETs (carcinoid tumour, including bronchial location; islet cell carcinoma; or other moderately or welldifferentiated tumours Locally advanced or metastatic, well or moderately differentiated NETs with clear evidence of pancreatic origin Carcinoid or pancreatic endocrine tumours that were not candidates for curative surgery Metastatic NETs of the digestive system of any grade with or without disease progression Metastatic or locally advanced G1 to G2 carcinoid tumours or pnets Well- to moderately differentiated, advanced (unresectable or metastatic) pnets mg oral sorafenib twice daily on days 1-5 every wk + 5mg/kg bevacizumab intravenously once every 2 wks for 6 months 58 25mg temsirolimus intravenously on days 1, 8, 15, mg/kg bevacizumab intravenously on days 1 and 15 of a 28-day cycle C=41 50mg oral sunitinib daily for 4 wks followed by 2 wks off treatment in repeated 6 wk cycles P=66 Dose escalation to 62.5mg and 75mg was permitted but not required µg pasireotide subcutaneously twice daily Dose increase as needed every 3 days for symptom control, up to a maximum of 900µg C,20 800mg pazopanib hydrochloride orally once per day for 12 treatment cycles, 28 days each + 30mg octreotide P,32 intramuscularly every 28 days. Patients already receiving depot octreotide continued receiving it at their preprotocol dose without exceeding 40mg every 3 wks 115 Stratum 1: 10mg oral everolimus daily OR Yao JC (2008) [48] Phase II Metastatic or unresectable locoregional, low- to intermediate-grade NETs Stratum 2: 10mg oral everolimus daily + 30mg octreotide LAR intramuscularly every 28 days 30mg octreotide LAR intramuscularly every 28 days + 5mg oral everolimus daily OR Retrospective studies (13 studies) Abbasi S (2014) [49] 30 30mg octreotide LAR intramuscularly every 28 days + 10mg oral everolimus daily Metastatic well differentiated NETs mg/m 2 capecitabine twice daily on days mg/m 2 temozolomide (increased to 200mg/m 2 in cycle 2 and beyond if well tolerated) divided into twice-a-day dosing on days of a 28-day cycle Section 4: Systematic Review - December 9, 2016 Page 17

21 Reference Inclusion criteria N Treatment Capdevila J (2015) [50] NETs 133 Lanreotide + novel targeted therapy Dilz LM (2015) [51] Dussol AS (2015) [52] Iwasa S (2010) [53] Kamp K (2013) [54] Khan MS (2011) [55] Koch W (2014) [17] Mirvis E (2014) [56] Palazzo M (2013) [57] Panzuto F (2014) [7] Ramundo V (2014) [58] Strosberg JR (2011) [59] Locally advanced or metastatic pnet mg/m 2 streptozocin mg/m 2 5-fluorouracil intravenously once daily on days 1-5 Advanced, well-differentiated NETs 104 GEMOX regimen: 1000mg/m 2 gemcitabine + 100mg/m 2 oxaliplatin every 2 wks; stopped at the end of 8 cycles Poorly differentiated neuroendocrine carcinomas arising 21 80mg/m 2 cisplatin intravenously over 2hrs on day mg/m 2 from the hepatobiliary tract and pancreas intravenously over 2hrs every day on days 1-3, repeated every 3-4 Well-differentiated or moderately differentiated (G1 or G2) GEP NETs that received any dose of everolimus following 177 Lu-octreotate Malignant carcinoid syndrome (metastatic NETs of midgut origin & symptoms of carcinoid syndrome) wks for a maximum of 6 cycles 24 5mg or 10mg everolimus following 177 Lu-octreotate 69 Deep subcutaneous injection with PR lanreotide with initial doses of 60, 90, or 120mg Well-differentiated (G1 or G2) metastatic NET of the ileum 30 Octreotide acetate 30 treatment according to ENETS guidelines Metastatic well-differentiated NETs 35 Starting dose of 1.5 to 3.5mU interferon-alpha three times per week Well-differentiated digestive neuroendocrine carcinomas 68 Lanreotide MP, lanreotide Autogel or lanreotide MP + lanreotide unsuitable for surgical resection with curative intent Autogel in succession Well- or moderately differentiated NETs representing mg everolimus daily unresectable or metastatic disease PI could start at or reduce dose to 5mg daily Multiple endocrine neoplasia type 1 patients with earlystage 20 30mg octreotide LAR intramuscularly every 28 days as first line duodeno-pancreatic NETs (<20mm in size) therapy Metastatic pancreatic endocrine carcinomas mg/m 2 oral capecitabine twice daily on days mg/m 2 oral temozolomide once daily at bedtime on days every 28 days 8mg ondansetron was prescribed before each dose of temozolomide Abbreviations: C, carcinoid tumour; CT, computed tomography; ENETS, European Neuroendocrine Tumour Society; G, grade; GEP, gastroenteropancreatic; hr(s), hour(s); LAR, long-acting repeatable; min(s), minute(s); MP, microparticles; MRI, magnetic resonance imaging; mtor, mammalian target of rapamycin; N, number of patients enrolled; NET, neuroendocrine tumour; P, pancreatic tumour; PI, principal investigator; pnet, pancreatic neuroendocrine tumour; PR, prolonged release; wk(s), week(s) Section 4: Systematic Review - December 9, 2016 Page 18

22 Study Design and Quality Quality of Individual Trials Quality characteristics were assessed for each of the 11 RCTs and are reported in Appendix 4 [1-4,6,14,15,18,31-33]. Four trials are currently only available in abstract form [4,6,31,32]. For trials with multiple published reports, all reports were searched for the necessary information. Allocation concealment was not reported in five studies [2,4,6,31,32]. Blinding was performed in seven trials [1,3,11,14,15,18,33]. Power and required sample size were calculated and reported in seven trials [1,3,6,11,14,15,18]. One trial had a loss to follow-up of patients [15] while four trials did not report on this feature [4,6,31,32]. Three of the RCTs were terminated early [2,32,33]; one because survival results favoured the treatment group [2]; another due to the drug safety profile [32]; and the last due to a low predictive probability of one arm showing superiority over the other for the primary outcome [33]. Furthermore, in three trials, participants assigned to the placebo arm could cross over to the treatment arm once disease progression was established by the investigator [2,3,18]; in another trial crossover was not allowed prior to primary analysis [14]; and in another trial patients who were not benefiting from octreotide by the end of the study core phase were able to cross over [33]. The 16 non-randomized prospective studies and 12 retrospective studies have critical risk of bias as a result of biases inherent in single-arm and retrospective studies. Quality of Aggregate Evidence According to GRADE, observational studies without special strengths or important limitations provide low-quality evidence. Case series, where there is no control group, usually result in downgrading of the evidence from low quality to very low quality. Pancreatic neuroendocrine tumours The quality of the aggregate evidence for chemotherapy is very low and was marked down due to risk of bias (prospective and retrospective single-arm studies) and imprecision (low patient numbers). The best evidence for targeted therapy comes from two RCTs [2,3], which were used to assess the overall quality. The quality of the evidence is moderate, marked down for risk of bias due to patients being able to cross over to the treatment arm at disease progression in both trials. The quality of the evidence for SSAs is from one RCT [1] and was found to be low due to imprecision (wide CI) and risk of bias (subgroup not powered for analysis). The best evidence for combination therapy comes from one RCT [4], which was used to assess the overall quality. The quality of the evidence is moderate due to imprecision (wide CI). Non-pancreatic neuroendocrine tumours The best evidence for targeted therapy comes from one RCT [29], which was used to assess the overall quality and was considered to be moderate, marked down for risk of bias (subgroup analysis not powered for analysis). The best evidence for SSAs comes from two RCTs [1,16], which were used to assess the quality and were considered to be low, marked down for imprecision (low patient numbers) and risk of bias (subgroup not powered for analysis). The quality of the aggregate evidence for combination therapy was very low as all trials were single-arm and were marked down due to indirectness, risk of bias (prospective single-arm studies), and imprecision (low patient numbers). Section 4: Systematic Review - December 9, 2016 Page 19