University of Groningen Neuroendocrine tumors; measures to ...Lotte D. de Hosson. Neuroendocrine...

University of Groningen

Neuroendocrine tumors; measures to improve treatment and supportive carede Hosson, Lotte Doortje

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2019

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):de Hosson, L. D. (2019). Neuroendocrine tumors; measures to improve treatment and supportive care.[Groningen]: Rijksuniversiteit Groningen.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 11-10-2020

https://www.rug.nl/research/portal/en/publications/neuroendocrine-tumors-measures-to-improve-treatment-and-supportive-care(c3938e28-5712-48b5-995d-7013808213ff).html

https://www.rug.nl/research/portal/en/persons/lotte-hosson-de(d6b0f3b8-94a5-44bd-8e31-5d08347f0cf7).html

https://www.rug.nl/research/portal/en/publications/neuroendocrine-tumors-measures-to-improve-treatment-and-supportive-care(c3938e28-5712-48b5-995d-7013808213ff).html

Neuroendocrine tumors;

Measures to improve treatment and supportive care.

Lotte D. de Hosson

Neuroendocrine tumors; Measures to improve treatment and supportive careLotte D. de Hosson

PhD dissertation, University of Groningen, the Netherlands

Layout by: XML 2 PublishPrinting by: Gildeprint, Enschede

ISBN: 978-94-034-1443-0ISBN (electronic version): 978-94-034-1442-3

© Copyright 2019 – Lotte D. de Hosson, Groningen, the Netherlands

All right reserved. No part of this thesis may be reproduced, stored in a retrieval or transmitted in any form or by any means, without prior permission of the author or, when appropriate of the publisher of the published articles. Printing of this thesis was financially supported by the Stichting Werkgroep Interne Oncologie, Rijksuniversiteit Groningen, Universitair Medisch Centrum Groningen.

Neuroendocrine tumors; Measures to improve treatment and

supportive care

Proefschrift

ter verkrijging van de graad van doctor aan de Rijksuniversiteit Groningen

op gezag van de rector magnificus prof. dr. E. Sterken

en volgens besluit van het College voor Promoties.

De openbare verdediging zal plaatsvinden op

woensdag 20 februari 2019 om 16.15 uur

door

Lotte Doortje de Hosson

geboren op 15 januari 1985 te Maurik

Promotor Prof. dr. E.G.E. de Vries

Copromotor Dr. A.M.E. Walenkamp

Beoordelingscommissie Prof. dr. W.W. Herder

Prof. dr. H. Hollema

Prof. dr. G.A.P. Hospers

Paranimfen Kim J.H.A. de Hosson Iris I.C. Kok

ContentsChapter 1 General introduction and outline of the thesis 6

Chapter 2 Clinical benefit of systemic treatment in patients with advanced pancreatic and gastro-intestinal neuroendocrine tumors according to ESMO-MCBS and the Net Health Benefit of ASCO revised framework.

16

L.D. de Hosson, L.M. van Veenendaal, Y. Schuller, W.T. Zandee, W.W. de Herder, M.E.T. Tesselaar, H.J. Klümpen, A.M.E. Walenkamp.

Chapter 3 a) Web-based tailored information and support for patients with a Neuroendocrine tumor; WIN-pilot.

40

G. Bouma, L.D. de Hosson, C. E. van Woerkom, H. van Essen, G.H. de Bock, J.M. Admiraal, A.K. L. Reyners, A. M. E. Walenkamp.

b) Web-based tailored nformation and support for patients with a Neuroendocrine tumor; WIN-study

58

L.D. de Hosson*, G. Bouma*, J. Stelwagen, H. van Essen, G.H. de Bock, D.J.A. de Groot, E.G.E. de Vries, A.M.E. Walenkamp.

Chapter 4 Towards optimal personalized diet and vitamin supplementation in NET patients; DIVIT-pilot.

78

L.D. de Hosson, J. Stelwagen, G. Bouma, B. Sijtema, S, Huitema, H.J.R. van Faassen, G.H. de Bock, D.J.A. de Groot, M.J.E. Campmans-Kuijpers, I.P. Kema, E.G.E. de Vries, A.M.E. Walenkamp.

Chapter 5 Heterogeneity in neuroendocrine tumor lesions visible with [18F]FDOPA PET/CT

88

L.D. de Hosson, A.M. van der Loo- van der Schaaf, R. Boellaard, J. H. van Snick, E.G.E. de Vries, A. H. Brouwers, A.M.E. Walenkamp.

Chapter 6 Reasons for the cold immune microenvironment of neuroendocrine tumours.

108

L.D. de Hosson*, G. Bouma*, G. Kats-Ugurlu, M. Bulthuis, E.G.E. de Vries, M. van Faassen, I.P. Kema, A.M.E. Walenkamp.

Chapter 7 Summary and future perspectives 126

Appendix Dutch summary, Nederlandse samenvatting 143

Appendix Dankwoord 151

1

General introduction and outline of the thesis

1

9


Background

Neuroendocrine tumors (NETs) are rare tumors with an incidence 3.5/100,000 per year and a prevalence of 35/100,000 (1-3). The tumor arises from secretory cells of the neuroendocrine system. Patients with NET may experience various symptoms from the tumor mass, the output of hormones secreted by the tumor, and treatment accompany-ing side effects (4). Median survival in patients with regional and distant metastatic NET is 105 and 56 months respectively, illustrating the indolent nature of most NETs. Patients with localized and metastatic gastro-intestinal NET who have lived with NET for 5 years have a probability of surviving an additional 5 years of 98.2% and 73.5% respectively (3,5). This underscores the fact that most patients live relatively long. Patients with NET have a lower quality of life (QoL) compared with the general population (6-8). Patients are frequently metastasized at the time of diagnosis. Radical surgery is the only curative treatment for patients presenting with local or regional disease, solitary metastases or resectable liver metastases. Non-curative systemic treatment options have variable, limited success (9-20). Ideally, in trials investigating novel drugs, the primary aim should be to study clinical relevant endpoints such as overall survival (OS) and QoL (21). However, in gastrointes-tinal and pancreatic NET (GEP-NET) patients, survival analysis can be challenging as the prolonged (natural) course of the disease often results in trials that allow crossover towards the experimental arm, or that allow second line therapies which could influence OS (22-24). Therefore, use of OS as primary endpoint is challenging, and QoL becomes more important to determine meaningful clinical benefit in the treatment of GEP-NET. In cancer survivors adequate information provision is associated with health related QoL and is an essential aspect of supportive care (25,26). For patients with NET it is difficult to find meaningful and understandable information about their diagnosis. Other important factors influencing QoL in patients with NET are gastro-intestinal problems and increased frequency of bowel movements (27). Furthermore somatostatin analogues (SSA) decrease their pancreatic exocrine function and can result in fat-soluble vitamin deficiencies (28-30). In case of high serotonin production, tryptophan deficiency, could also lead to niacin (vitamin B3) deficiency and lead to symptoms (29,31). Currently, strik-ingly little is known about how SSA using patients with NET should be best supported and if dietary support and supplementation of vitamins is feasible (32,33).Clinical symptoms are, among others, associated with reduction in QoL. Symptoms in patients with NET can be caused by the high secretion of neuroendocrine amines. Therefore, local treatment of an active amine producing tumor lesion would be poten-tially meaningful to improve QoL (34). Previously it was demonstrated that tumor burden measured with total [18F]FDOPA uptake correlates with catecholamine pathway activity (35). If lesions with high [18F]FDOPA uptake could be identified with a non-invasive

Chapter 1

10

method in patients with small intestinal NET, this might be of added value in the selec-tion of lesions for local treatment. Over the last years, immunotherapy with immune checkpoint inhibitors like the pro-grammed death-1 and programmed death ligand-1 (PD1 and PD-L1) antibodies have shown antitumor activity across numerous tumor types by activation of T-cells in the tumor microenvironment (36). Interestingly for NETs, there is also a major interest in the tryptophan-degrading enzymes indoleamine 2,3-dioxygenase (IDO) and tryptophan 2,3-dioxygenase (TDO) which are involved in the immune response and are expressed in several tumor types. In serotonin producing NETs IDO and TDO play a special im-portant role since tryptophan is the precursor of the serotonin pathway (37,38). There is, however limited information with regards to the complex interactions of NET tumour cells with their surrounding immune microenvironment. Consequently knowledge about potential targets for immunotherapy in patients with NET is limited.

Aim of the thesis

The aim of this thesis is to measure and improve the clinical benefit of treatment and supportive care provided to patients with NET. We investigated tailored supportive measures for these patients as well as novel treatment approaches.

Outline of the thesis

In chapter 2 an overview of existing literature about trials comparing systemic antitumor treatment for GEP-NET versus a control is presented and their clinical benefit is inter-preted. The electronic databases (PubMed, Embase) were searched for papers report-ing comparative trials, conducted in adult patients with GEP-NET. The clinical benefit of the investigated therapies reported in the selected trials was determined according to the European Society of Medical Oncology Magnitude of Clinical Benefit Scale (ESMO-MCBS) and the net health benefit score according to the American Society of Clinical Oncology (ASCO) revised framework by four independent assessors.QoL is an important factor that contributes to clinical benefit. Factors associated with enhanced QoL include; less clinical symptoms, and adequate information provision (39,40). The internet was found to be a useful source of general information for some patients with NET (41). In chapter 3A the feasibility of a web based system consist-ing of self-screening of problems and care needs in patients with NET was analyzed. Newly diagnosed NET patients were randomized between standard care (n=10) or intervention with additional access to the web-based system (n=10) during 12 weeks.

1

11


Patients completed questionnaires regarding received information, distress, QoL, and empowerment. The intervention group completed a semi-structured interview to assess patients’ opinion on the web-based system.Subsequently, in a randomized controlled trial 90 patients with NET were stratified between those newly diagnosed (< 6 months, n=28) or with a longer duration of the disease (n=74) and randomized between standard care (n=49) or intervention with ad-ditional access to the web-based system (n=53) during 12 weeks. Patients completed questionnaires about distress, perception and satisfaction of received information, QoL, and empowerment. The intervention group also completed a questionnaire based on the technical acceptance model regarding their use of and opinion on the web-based system (chapter 3B). In chapter 4 we described a feasibility study, in which 15 patients with NET using a SSA for over 6 months were counseled by a dietician for a personalized dietary advice and received supplementation of deficient vitamins A, D, E, K, B12 and vitamin B3 (niacin). Feasibility was assessed by calculation of participation/dropout rate and number of (severe) adverse events related to the intervention. At baseline, after 4 and 18 weeks, vitamins were measured, and QoL, distress, empowerment and nutrition state were assessed.Part of symptoms in patients with NET are caused by the secretion of neuroendocrine amines and local treatment of a high metabolic active lesion would be another potential method to improve QoL meaningful in patients with NET (34). Therefore, in chapter 5 we analyzed intertumoral heterogeneity with 18F-dihydroxyphenylalanine (DOPA) Posi-tron Emission Tomography (PET) scans in patients with small intestinal (SI-) NET and investigated if tumor lesions with substantially higher [18F]FDOPA uptake than the ma-jority of the other lesions within a patient could be identified within a patient. [18F]FDOPA PET scans of 38 patients, of which 35 serotonin producing, were analyzed. For all tumor lesions the [18F]FDOPA uptake was calculated by dividing the standard uptake value (SUV) peak of the tumor lesion by the SUV mean of the background organ. The magnitude of heterogeneity between lesions within a patient was calculated by dividing the lesion with highest [18F]FDOPA uptake by the one with lowest [18F]FDOPA uptake. In patients (n=20) with more than 10 metastases the lesions with a higher tracer uptake than the upper inner or outer fence (more than 1.5 or 3 times the interquartile range above the third quartile) were defined as lesions with mild or extreme high [18F]FDOPA uptake respectively. In chapter 6 we investigated the immune microenvironment of neuroendocrine tumors, by pathologic analysis of tumor tissue of 51 patients with serotonin or non-serotonin producing NET grade 1 or 2. Immunohistochemically analyses were performed for PD-L1, T-cells, IDO, TDO, mismatch repair proteins and activated fibroblasts. Serotonin

Chapter 1

12

production was measured by high performance liquid chromatography fluorometry of 5-hydroxyindolacetic acid (5-HIAA) in 24-h urine and/or serotonin in platelet rich plasma.Finally the findings of this thesis are summarized in chapter 7 followed by a discussion on the scientific and clinical implications and direction for future research. Chapter 8 comprises the summary of this thesis in Dutch.

1

13


References 1. Lawrence B, Gustafsson BI, Chan A, Svejda B, Kidd M, Modlin IM. The epidemiology of gas-

troenteropancreatic neuroendocrine tumors. Endocrinol Metab Clin North Am 2011;40:1-18. 2. Dasari A, Shen C, Halperin D, et al. Trends in the incidence, prevalence, and sur-

vival outcomes in patients with neuroendocrine tumors in the United States. JAMA Oncol 2017;3:1335-42.

3. Yao JC, Hassan M, Phan A, et al. One hundred years after “carcinoid”: epidemiology of and prognostic factors for neuroendocrine tumors in 35,825 cases in the United States. J Clin Oncol 2008;26:3063-72.

4. Oberg KE. The management of neuroendocrine tumours: current and future medical therapy options. Clin Oncol (R Coll Radiol) 2012;24:282-93.

5. Mocellin S, Nitti D. Gastrointestinal carcinoid: epidemiological and survival evidence from a large population-based study (n = 25 531). Ann Oncol 2013;24:3040-4.

6. Beaumont JL, Cella D, Phan AT, Choi S, Liu Z, Yao JC. Comparison of health-related quality of life in patients with neuroendocrine tumors with quality of life in the general US popula-tion. Pancreas 2012;41:461-6.

7. Frojd C, Larsson G, Lampic C, von Essen L. Health related quality of life and psychosocial function among patients with carcinoid tumours. A longitudinal, prospective, and compara-tive study. Health Qual Life Outcomes 2007;5:18.

8. Haugland T, Vatn MH, Veenstra M, Wahl AK, Natvig GK. Health related quality of life in patients with neuroendocrine tumors compared with the general Norwegian population. Qual Life Res 2009;18:719-26.

9. Raymond E, Dahan L, Raoul JL, et al. Sunitinib malate for the treatment of pancreatic neuroendocrine tumors. N Engl J Med 2011;364:501-13.

10. Faivre S, Niccoli P, Raoul JL, et al. Updated overall survival (OS) analysis from a phase III study of sunitinib vs placebo in patients (PTS) with advanced, unresectable pancreatic neuroendocrine tumor (NET). Ann Oncol 2012;23:ix376.

11. Rinke A, Muller HH, Schade-Brittinger C, et al. Placebo-controlled, double-blind, prospec-tive, randomized study on the effect of octreotide LAR in the control of tumor growth in patients with metastatic neuroendocrine midgut tumors: a report from the PROMID Study Group. J Clin Oncol 2009;27:4656-63.

12. Rinke A, Wittenberg M, Schade-Brittinger C, et al. Placebo-controlled, double-blind, pro-spective, randomized study on the effect of octreotide LAR in the control of tumor growth in patients with metastatic neuroendocrine midgut tumors (PROMID): Results of long-term survival. Neuroendocrinology 2017;104:26-32.

13. Caplin ME, Pavel M, Cwikla JB, et al. Lanreotide in metastatic enteropancreatic neuroen-docrine tumors. N Engl J Med 2014;371:224-33.

14. Yao JC, Shah MH, Ito T, et al. Everolimus for advanced pancreatic neuroendocrine tumors. N Engl J Med 2011;364:514-23.

15. Yao JC, Pavel M, Lombard-Bohas C, et al. Everolimus for the treatment of advanced pancreatic neuroendocrine tumors: Overall survival and circulating biomarkers from the randomized, phase III RADIANT-3 study. J Clin Oncol 2016;34:3906-13.

16. Yao JC, Fazio N, Singh S, et al. Everolimus for the treatment of advanced, non-functional neuroendocrine tumours of the lung or gastrointestinal tract (RADIANT-4): a randomised, placebo-controlled, phase 3 study. Lancet 2016;387:968-77.

Chapter 1

14

17. Meyer T, Qian W, Caplin ME, et al. Capecitabine and streptozocin +/- cisplatin in advanced gastroenteropancreatic neuroendocrine tumours. Eur J Cancer 2014;50:902-11.

18. Pavel ME, Hainsworth JD, Baudin E, et al. Everolimus plus octreotide long-acting repeatable for the treatment of advanced neuroendocrine tumours associated with carcinoid syndrome (RADIANT-2): a randomised, placebo-controlled, phase 3 study. Lancet 2011;378:2005-12.

19. Pavel M, Oberg KE, Hainsworth JD, et al. Everolimus plus octreotide long-acting release (LAR) for the treatment of advanced neuroendocrine tumors (NET) associated with carci-noid syndrome (RADIANT-2): Updated overall survival results. Eur J Cancer 2013;49:S577.

20. Strosberg J, El-Haddad G, Wolin E, et al. Phase 3 trial of 177Lu-dotatate for midgut neuro-endocrine tumors. N Engl J Med 2017;376:125-35.

21. Booth CM, Eisenhauer EA. Progression-free survival: meaningful or simply measurable? J Clin Oncol 2012;30:1030-3.

22. Kulke MH, Siu LL, Tepper JE, et al. Future directions in the treatment of neuroendocrine tumors: consensus report of the National Cancer Institute Neuroendocrine Tumor clinical trials planning meeting. J Clin Oncol 2011;29:934-43.

23. Lebwohl D, Kay A, Berg W, Baladi JF, Zheng J. Progression-free survival: gaining on overall survival as a gold standard and accelerating drug development. Cancer J 2009;15:386-94.

24. Riechelmann RP, Rego J. Progression-free survival in neuroendocrine tumors: preferred end point, but how should it be defined? J Clin Oncol 2011;29:2835-6.

25. Rehnberg G, Absetz P, Aro AR. Women’s satisfaction with information at breast biopsy in breast cancer screening. Patient Educ Couns 2001;42:1-8.

26. Husson O, Mols F, van de Poll-Franse LV. The relation between information provision and health-related quality of life, anxiety and depression among cancer survivors: a systematic review. Ann Oncol 2011;22:761-72.

27. Pearman TP, Beaumont JL, Cella D, Neary MP, Yao J. Health-related quality of life in pa-tients with neuroendocrine tumors: an investigation of treatment type, disease status, and symptom burden. Support Care Cancer 2016;24:3695-703.

28. Fiebrich HB, Van Den Berg G, Kema IP, et al. Deficiencies in fat-soluble vitamins in long-term users of somatostatin analogue. Aliment Pharmacol Ther 2010;32:1398-404.

29. Shah GM, Shah RG, Veillette H, Kirkland JB, Pasieka JL, Warner RR. Biochemical as-sessment of niacin deficiency among carcinoid cancer patients. Am J Gastroenterol 2005;100:2307-14.

30. Lembcke B, Creutzfeldt W, Schleser S, Ebert R, Shaw C, Koop I. Effect of the somatostatin analogue sandostatin (SMS 201-995) on gastrointestinal, pancreatic and biliary function and hormone release in normal men. Digestion 1987;36:108-24.

31. Bouma G, van Faassen M, Kats-Ugurlu G, de Vries EG, Kema IP, Walenkamp AM. Niacin (Vitamin B3) supplementation in patients with serotonin-producing neuroendocrine tumor. Neuroendocrinology 2016;103:489-94.

32. Maton PN. The carcinoid syndrome. JAMA 1988;260:1602-5. 33. Mohyi D, Tabassi K, Simon J. Differential diagnosis of hot flashes. Maturitas 1997;27:203-

14. 34. Pavel M, Baudin E, Couvelard A, et al. ENETS Consensus Guidelines for the manage-

ment of patients with liver and other distant metastases from neuroendocrine neoplasms of foregut, midgut, hindgut, and unknown primary. Neuroendocrinology 2012;95:157-76.

1

15


35. Fiebrich HB, de Jong JR, Kema IP, et al. Total 18F-dopa PET tumour uptake reflects meta-bolic endocrine tumour activity in patients with a carcinoid tumour. Eur J Nucl Med Mol Imaging 2011;38:1854-61.

36. Guan J, Lim KS, Mekhail T, Chang CC. Programmed death ligand-1 (PD-L1) expression in the programmed death receptor-1 (PD-1)/PD-L1 blockade: A key player against various cancers. Arch Pathol Lab Med 2017;141:851-61.

37. Uyttenhove C, Pilotte L, Theate I, et al. Evidence for a tumoral immune resistance mechanism based on tryptophan degradation by indoleamine 2,3-dioxygenase. Nat Med 2003;9:1269-74.

38. van Baren N, Van den Eynde BJ. Tryptophan-degrading enzymes in tumoral immune resis-tance. Front Immunol 2015;6:34.

39. Davies AH, Larsson G, Ardill J, et al. Development of a disease-specific Quality of Life questionnaire module for patients with gastrointestinal neuroendocrine tumours. Eur J Cancer 2006;42:477-84.

40. Yadegarfar G, Friend L, Jones L, et al. Validation of the EORTC QLQ-GINET21 question-naire for assessing quality of life of patients with gastrointestinal neuroendocrine tumours. Br J Cancer 2013;108:301-10.

41. Feinberg Y, Law C, Singh S, Wright FC. Patient experiences of having a neuroendocrine tumour: a qualitative study. Eur J Oncol Nurs 2013;17:541-5.

2

Clinical benefit of systemic treatment in patients with advanced

pancreatic and gastro intestinal neuroendocrine tumors according to ESMO-MCBS and ASCO framework

L.D. de Hosson1, L.M. van Veenendaal2, Y. Schuller3, W.T. Zandee4, W.W. de Herder4, M.E.T. Tesselaar2, H.J. Klümpen3, A.M.E. Walenkamp1

1Department of Medical Oncology, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands

2Department of Medical Oncology, Netherlands Cancer Institute, Amsterdam, The Netherlands

3Department of Medical Oncology, Academic Medical Center, Amsterdam, The Netherlands

4Department of Endocrinology, Erasmus Medical Center, Rotterdam, The Netherlands

Chapter 2

18

Abstract:

BackgroundAssessment of clinical benefit of systemic treatments of rare diseases including gastro-enteropancreatic neuroendocrine tumors (GEP-NET) is challenging. Recently several tools have been developed to grade clinical benefit of cancer drugs. European Society for Medical Oncology (ESMO) has developed the ESMO Magnitude of Clinical Benefit Scale (ESMO-MCBS). The American Society of Clinical Oncology (ASCO) developed and revised the ASCO framework consisting of the Net Health Benefit (NHB) score juxtaposed against the costs of the treatment. In this review we graded systemic treat-ments for GEP-NET patients with both frameworks.

MethodsThe electronic databases (PubMed, Embase) were searched for papers reporting comparative trials, conducted in adult GEP-NET patients in the English language. Papers were assessed according to the ESMO-MCBS and the NHB part of the ASCO revised Framework (NHB-ASCO-F) by 4 independent assessors, discrepancies were discussed.

ResultsThe search yielded 32 trials of which 6 trials were eligible for grading with the ESMO-MCBS resulting in scores of 2 or 3. Eight trials were eligible for grading with the NHB-ASCO-F; resulting in scores between 37.6 and 57.4. Trials that were not primary assessable by the tools were analyzed separately. Consensus between assessors was reached in 68% of trials with the ESMO-MCBS and in 23% of trials with the NHB-ASCO-F.

ConclusionThe currently used systemic treatments for GEP-NET patients had low scores accord-ing to the NHB-ASCO-F and none could be graded as meaningful clinical beneficial according to the ESMO-MCBS. Despite the low incidence, the heterogeneous patient population and relatively long natural course of NET, future studies on new treatment modalities should aim for high clinical benefit outcomes. Keywords: ESMO-MCBS, ASCO, pancreatic and gastro-intestinal, neuroendocrine tumors, clinical benefit, value

2

19

Clinical benefit of systemic treatment in patients with advanced pancreatic and gastro intestinal neuroendocrine tumors according to ESMO-MCBS and ASCO framework

Introduction

Neuroendocrine tumors are rare malignancies with an incidence of 3.5/100,000 per year and a prevalence of 21.6/100,000 in the last decade (1). Median survival time is 77 months in patients with regional and 24 months in patients with distant metastatic pancreatic NET (pNET) and 105 months and 56 months in case of respectively regional and distant metastatic intestinal NET disease (2). GEP-NET patients are frequently metastasized at the time of initial diagnosis. Surgery is the only curative treatment. Non-curative systemic treatment options include somatostatin analogues (SSA), che-motherapy, targeted agents, and peptide receptor radionuclide therapy (PRRT) with variable, limited success (3-12).Ideally, in clinical trials investigating novel drugs, the primary aim should be to study endpoints such as the overall survival (OS), quality of life (QoL) and treatment toxicity. However, for trials investigating new agents in GEP-NET patients, survival analysis can be challenging as the prolonged (natural) course of the disease often results in trials al-lowing crossover towards the experimental arm, which influences OS. Crossover to the experimental arm or to second line therapies has impact on OS data (13). Therefore, progression-free survival (PFS) was recommended as a feasible and relevant primary end point for both, phase II and III trials in GEP-NET, by the expert consensus report of the National Cancer Institute Neuroendocrine Tumor Clinical Trials Planning Meeting (13).Recently, among others, ESMO has developed a validated and reproducible tool to assess the magnitude of clinical benefit for drugs for solid tumors, the ESMO-MCBS (14). The ASCO developed the ASCO framework for solid tumors and haematological malignancies consisting of the NHB score juxtaposed against the costs of the treatment (15,16). Both tools have been applied in several solid malignancies (14-18). However, the ASCO framework was assessed only in a few studies. Treatments for GEP-NET are not yet evaluated using the tools. In this review we therefore investigated the value of current systemic antitumor treatments for GEP-NET patients and their eligibility for grading with both ESMO-MCBS and NHB-ASCO-F.

Methods

Search strategyFor detailed description of methods see supplementary data. In brief, trials were searched between Aug 1st 2015 to Jan 31th 2016, in the databases PubMed and EM-BASE. The articles that were found, were screened using title and abstract to select trials published in the English language and comparing systemic treatment modalities

Chapter 2

20

for the treatment of GEP-NET in humans. Furthermore the reference lists of national and international guidelines, included trials and conference abstracts were reviewed for additional relevant articles.

Selection criteria for trialsComparative trials investigating systemic antitumor treatment for well-differentiated GEP-NET patients were analyzed. Studies in patients with grade 3 neuroendocrine neoplasms were not included. Criteria for assessment are summarized in Table S1. Tri-als were selected if at least 50 % of participants were diagnosed with GEP-NET, and the other participants had a NET of unknown or other origin. Abstracts were not included.

ESMO-MCBS and the NHB-ASCO-FESMO-MCBS grades, in the non-curative setting, range from 1-5, with grade 4 and 5 representing meaningful clinical benefit (14). The NHB-ASCO-F ranges from -20 to 180, with a higher score representing a better score, no cut-off value was provided to define clinical benefit (15,16,18). All relevant comparative trials were assessed according to both tools. Relevant trials that did not meet all criteria as mentioned in Table S1 were assessed separately, based on the available data.

Assessment of trialsFour members from all Dutch European Neuroendocrine Tumor Society (ENETS) cen-tres of excellence independently scored the trials according to the ESMO-MCBS and the NHB-ASCO-F. After obtaining scores of these assessors we noticed a wide variation in results. During a consensus meeting additional agreements were defined. Next, trials were assessed again according to the additional agreements. The number of trials to which the same score was awarded between three or four assessors was registered after each of the two scoring sessions. Two phase III trials were published after the consensus meeting and were assessed according our agreements (8,12). Therefore an additional assessment session was not necessary for these trials.

Results

Trial selection and characterization The primary search strategy yielded 1,942 potentially relevant papers, of which 1,676 remained when duplicates were discarded (Figure S1). After screening of title and abstract, 223 papers were preselected evaluating systemic antitumor treatment for GEP-NET patients.

2

21


Thirty nine papers in which 35 comparative trials in GEP-NET patients were described were selected for further analysis. Nine trials fulfilled all criteria as summarized in Table S1 and were assessable according to the ESMO-MCBS and the NHB-ASCO-F (Table 1) (3-12). Only one trial, the RADIANT-4, described in their protocol that crossover was not allowed (10,11). None of the trials investigated adjuvant treatment in GEP-NET or had a curative intent. The other 26 comparative trials, investigating systemic treatment in GEP-NET patients, did not meet all criteria for assessment with both tools (Table S1), or evaluated partici-pants of which less than 50% had a GEP-NET. Main difficulties for assessment with the tools were summarized in Figure S2. As currently used national and international guidelines are based upon these trials, we analyzed them separately with the available data (Table S2, supplementary data; references).

Assessment according to the ESMO-MCBSSix trials were assessable with the ESMO-MCBS. The calculated score ranged from 2 to 3 corresponding with a low level of clinical benefit. RADIANT-3 and RADIANT-4 analyzed everolimus versus placebo in pNET and non-functional advanced NET, re-spectively (9-11). Application of ESMO-MCBS for RADIANT-3 and RADIANT-4 resulted in a preliminary score of 3 reflecting a longer PFS. A hazard ratio with a lower limit of the 95% confidence interval ≤ 0.65 for PFS determined the preliminary score. No significant difference in important adverse events as compared to the placebo control arm was detected. QoL of everolimus in pNET patients was reported in a recent abstract which described a single-arm phase IV study, performed in patients who started with evero-limus (19). After 6 months, no improvement in QoL was detected. Because this was a single-arm study and the ESMO-MCBS could only be applied to comparative outcome studies, the preliminary score of the RADIANT-3 will not be downgraded, despite lack of improvement of QoL, when these data are published in full-report. QoL of RADIANT-4 trial was reported in a post-hoc analysis and recently published (11). No improvement of QoL by everolimus versus the control arm was demonstrated. Therefore the preliminary score of the RADIANT-4 was downgraded to a final score 2. The NETTER-1 trial inves-tigated 177Lu-dotatate with octreotide LAR versus octreotide LAR alone in patients with metastatic midgut NET (8). The calculated ESMO-MCBS score was 3, which implies a longer PFS in the intervention group as compared with control arm. Furthermore, there was no significant difference in important adverse events as compared to the control arm and no QoL data were documented. In the CLARINET trial GEP-NET patients or patients with NET of unknown origin were treated with lanreotide in the intervention arm versus placebo in the control arm (3). This trial showed an improvement in PFS. Data of QoL did not show an improvement in QoL for the intervention group, resulting in a score of 2. In the PROMID trial patients with a metastasized midgut NET were

Chapter 2

22

2

23


Tabl

e 1:

Clin

ical

ben

efit o

f sys

tem

ic tr

eatm

ent f

or G

EP-N

ET a

ccor

ding

to E

SMO

-MC

BS

(ESM

O-M

CB

S) a

nd a

ccor

ding

ASC

O r

evis

ed fr

amew

ork

(NH

B)

ESM

O-M

CB

SA

SCO

revi

sed

fram

ewor

k

Inte

rven

tion

vs c

ontr

olSe

tting

and

refe

renc

ePr

imar

y O

utco

me

HR

of

prim

ary

outc

ome/

RR

Into

xica

tion

data

Impr

ovem

ent

QO

L (Y

es/N

o)ES

MO

- M

CB

SC

linic

al b

enefi

t sc

ore

Into

xica

tion

data

Bon

us

poin

tsN

et

heal

th

bene

fit

Lanr

eotid

e vs

pla

cebo

GEP

-NET

or N

ET o

f unk

now

n or

igin

(3)

PFS

0.47

(0

.30-

0.73

)N

SN

o2

42.4

NC

0N

A

Cap

ecita

bine

and

st

rept

ozoc

in a

nd c

ispl

atin

vs

cap

ecita

bine

and

st

rept

ozoc

in

Adva

nced

, irre

sect

able

, NET

of

panc

reas

, gas

trodu

oden

um, o

r un

know

n pr

imar

y (4

)

RR

16 %

vs

12 %

NS

No

NA

0N

C0

NA

Oct

reot

ide

and

ever

olim

us

vs o

ctre

otid

e an

d pl

aceb

oAd

vanc

ed, p

rogr

essi

ve N

ET w

ith

carc

inoi

d sy

ndro

me

(5)

PFS

0.77

(0.5

9-1.

00)

NS

NA

NA

18.4

NC

0N

A

Suni

tinib

vs

plac

ebo

Prog

ress

ive,

adv

ance

d pa

ncre

atic

N

ET (6

)PF

S0.

42

(0.2

6-0.

66)

NS

No

246

.4-5

1657

.4

Oct

reot

ide

LAR

vs

plac

ebo

Firs

t lin

e m

etas

tatiz

ed m

idgu

t N

ET (7

)PF

S0.

34

(0.2

0-0.

59)

NS

No

252

.8N

C16

NA

177L

u-D

otat

ate

vs

octr

eotid

e LA

RM

etas

tast

atic

mid

gut N

ET (8

)PF

S0.

21

(0.1

3-0.

33)

NS

NA

363

.8N

C16

NA

Ever

olim

us v

s pl

aceb

oAd

vanc

ed p

ancr

eatic

NET

(9)

PFS

0.34

(0

.26-

0.44

)N

SN

A3

52.8

-20

1648

.8

Ever

olim

us v

s pl

aceb

oAd

vanc

ed, p

rogr

essi

ve n

on-

func

tiona

l lun

g or

GI-N

ET (1

0,11

)PF

S0.

48

(0.3

5-0.

67)

NS

No

241

.6-2

016

37.6

Bev

aciz

umab

and

oc

treo

tide

LAR

vs

inte

rfer

on a

lfa-2

b an

d oc

treo

tide

LAR

Adva

nced

, NET

with

pro

gres

sion

or

oth

er p

oor p

rogn

ostic

feat

ures

(1

2)

RR

0.90

(0

.72-

1.12

)N

SN

AN

A8.

4N

C0

8.4

A: N

ame

of fi

rst a

utho

r. R

efer

ence

s ar

e re

porte

d in

refe

renc

e lis

t. AS

CO

= Am

eric

an S

ocie

ty o

f Clin

ical

Onc

olog

y, E

SMO

-MC

BS=

Euro

pean

Soc

iety

of M

edic

al O

ncol

ogy-

Mag

nitu

de o

f Clin

ical

Ben

efit S

cale

, GEP

-NET

=gas

troin

test

inal

and

/ or

panc

reat

ic n

euro

-end

ocrin

e tu

mor

, HR

= ha

zard

ratio

, NET

=neu

roen

docr

ine

tum

or, v

s=ve

rsus

, LAR

=lon

g ac

ting

reag

ents

, NA=

not

app

licab

le, N

C=

no c

onse

nsus

, NS=

no s

igni

fican

t di

ffere

nce.

OS=

over

all s

urvi

val.

PFS=

prog

ress

ion

free

surv

ival

. QoL

=qua

lity

of li

fe. R

R=r

espo

nse

rate

. SSA

=som

atos

tatin

ana

logu

e vs

= ve

rsus

.

Chapter 2

22

2

23


randomized between octreotide LAR and placebo (7). Further enrolment was stopped after inclusion of 85 patients instead of the planned 162 patients, because of observed positive effects of octreotide LAR on tumor growth and a slow recruitment rate. This trial resulted in an ESMO-MCBS score of 2. The trial reported by Raymond et al. analyzed sunitinib versus placebo in patients with progressive, advanced pancreatic NET (6). Ap-plication of ESMO-MCBS resulted in a score of 2. An improvement in PFS was shown. Because reported QoL data did not show improvement, the preliminary score of 3 had to be downgraded one point. This trial had an early closure after randomization of 171 patients, as observed and recommended by the safety monitoring board due to more serious adverse events and a higher frequency of death in the placebo group as well as a difference in PFS favoring sunitinib. The trial analyzing SSA and interferon versus SSA and bevacizumab, in advanced NET patients with progression or other poor prog-nostic features, showed no significant difference in its primary endpoint PFS (12). The RADIANT-2 analyzed everolimus with long acting octreotide in advanced NET tumor patients associated with carcinoid syndrome (5). The PFS with a hazard ratio of 0.77 (0.59-1.00) did not show a statistically significant clinical benefit. The trial analyzing capecitabine and streptozocin with or without cisplatin did also not show a statistically significant clinical benefit (4). Furthermore in these trials no improvement in toxicity, Qol, or OS was seen, and therefore the ESMO-MCBS was not applicable .Consensus in the ESMO-MCBS score was reached in 2 of 6 trials after the first scor-ing session (3-7,9-11). The variation in awarded scores was related to differences in interpretation of data, like the significance of the primary endpoint, and adjustment of the preliminary score for lack of improvement in QoL. After the second scoring session consensus was reached in 4 of 6 trials (3-7,9-11).

Assessment according to the NHB-ASCO-FNine trials were assessable for the NHB-ASCO-F (3-12). The CBS varied from 0 to 63.8. In the trial reported by Raymond et al. and the RADIANT-3 and RADIANT-4 trials the treatment arm experienced more adverse events, resulting in a toxicity score of -5 and -20, respectively (6,9-11). Bonus points were awarded for long term disease control; if, at a time point that was twice the median PFS for the control regimen, the percentage of patients having PFS was at least 50% higher for the intervention arm compared with the control arm, to the trial reported by Raymond et al, PROMID-trial, RADIANT-4, RADIANT-3 and the NETTER-1 (6-11). No consensus was obtained for the score of the trials assessed by the NHB-ASCO-F after 2 scoring sessions. This was generally related to discrepancies in interpretation of toxicity score. Finally, after an additional discussion in 4 trials consensus was obtained for NHB and in 5 trials consensus was reached for the CBS and bonus points.

Chapter 2

24

Discussion

To define the clinical benefit of systemic antitumor treatment in GEP-NET patients, we systematically applied the ESMO-MCBS and the NHB-ASCO-F to relevant trials. Six out of 35 trials fulfilled all requirements to be assessed with the ESMO-MCBS resulting in an ESMO-MCBS score of 2 or 3, while 9 trials could be assessed with the NHB-ASCO-F and resulted in scores between 37.6 to 57.4. No clear cut-off value to define clinical benefit with the ASCO framework was provided (15,16). None of the trials that were included in our analysis could demonstrate a meaningful clinical benefit according to the employed tools. The ESMO-MCBS scores were generally lower than in other tumor types, like metastasized breast or colorectal cancer (14,17). In more common tumor types, sufficient numbers of patients can be included in trials powered to detect a difference in OS between the intervention and control group. In NET, such large trials are scarce. In addition, OS difference detection is challenging in NET patients. There-fore, PFS is a frequently used, primary endpoint in NET trials. An improvement in PFS has less impact compared to OS in the ESMO-MCBS scores and the NHB-ASCO-F. Furthermore, none of the assessable trials in GEP-NET, showed an improvement in QoL or less toxicity as compared to the control arm. Therefore, these outcomes did not result in an upgrade of the preliminary score in the ESMO-MCBS. Ideally OS and QoL should be the primary endpoint of all trials with new drugs (20). This is challenging in NET. Reasons for this include the heterogeneous patient popu-lation, the long natural course of the disease and the wide variability of subsequent lines of therapy after progression. The heterogeneity of the patients composing the NET population is important to take into account. This illustrates the importance of well-defined inclusion criteria. The eligibility criteria of the RADIANT-2 trial included patients with; serotonin producing NET and progressive, low- or intermediate-grade and advanced disease (5). Despite these clear criteria still large variation in patient and tumor characteristics between patients exists. Despite the randomized design of the trial, this could have influenced PFS. Current guidelines for GEP-NET anti-tumor treatment are based on clinical trials and on expert opinion. These guidelines influence our daily clinical practice. Therefore, we also assessed trials that could not be fully assessed by the ESMO-MCBS and ASCO framework (table S2). The control arms of trials analyzing chemotherapy included interferon, dacarbazine, and other kinds of chemotherapy, respectively. Given current knowledge about the effects of chemotherapy, the control arm, nowadays, likely would not have contained chemotherapy. With this analysis we demonstrated, that current guidelines are partially based on studies not powered to determine endpoints that show clinical benefit.

2

25


Large trials were conducted by enormous efforts of the international NET community (3-12). This has provided us with important data. Improvements could include trials using a pre-selected patient population. This is already implemented in PRRT, where somatostatin-receptor-positive patients are selected for treatment (8). Furthermore, sequential multiple assignment randomized trials can be used for hypothesis genera-tion followed by a confirmatory randomized controlled trial (RCT) (21). In addition, trials should be focused on reporting QoL and be of sufficient follow-up duration. For GEP-NET patients, use of QoL could be of additional importance because OS is generally prolonged in this cohort.The importance of long follow-up was previously demonstrated in a trial in patients with anaplastic oligodendrogliomas. OS of adjuvant chemoradiotherapy versus radiotherapy was not improved at a median follow-up of 60 months, but was improved in the chemo-radiotherapy arm at a median follow-up of 140 months (22). International collaboration in clinical trials and investigator-driven trials should be further expanded to increase evidence based data to support treatment decisions and lead to meaningful clinical benefit for GEP-NET patients. Although not the aim of our analysis, we found some limitations of the tools. The ESMO-MCBS downgrades the preliminary score of a trial if no significant improvement of QoL data was shown (14). If QoL was not reported, the preliminary score did not have to be downgraded (8,9). Furthermore, we noticed during the scoring process, a difference in consensus achieved by assessing trials according to the ESMO-MCBS, compared to trials assessed ac-cording to the NHB-ASCO-F. With the ESMO-MCBS for more trials consensus was achieved. A possible reason for this is the complex toxicity data reporting according to the NHB-ASCO-F. Potentially, a clearer definition of ‘clinical relevant toxicity’ could facilitate the scoring. With the ASCO framework, eight clinicians completed the tool for 11 anticancer drugs. A Cohen’s kappa coefficient of the interrater reliability (ICC) of 0.11 for NHB and 0.06 for toxicity was found, corresponding with ´slightly reliable agreement’ (18).However in another study where convergent validity and interrater reliability of as-sessment frameworks were analyzed; the ASCO framework had an ICC of 0.80 (23). In another report, 109 RCTs were included and ESMO scores and scores using the concept and revised version of the ASCO framework were determined. Weak to moder-ate correlations were demonstrated, suggesting different constructs of clinical benefit measured (24). Other tools, used to define the value and also addressing the costs of drugs, include the National Comprehensive Cancer Network evidence blocks, Institute for Clinical and Economic Review value assessment framework and Memorial Sloan Kettering Drug Abacus (25-27). Every tool has its own aspects. In the future converging

Chapter 2

26

of tools is expected which will allow to combine the best aspects of these tools to guide policy makers and patient-doctor discussions (28).

Conclusion

The ESMO-MCBS and the NHB of ASCO revised framework could be applied, respec-tively, to only 6 and 9 trials investigating systemic treatments in NET patients. The cur-rently used systemic treatments for GEP-NET patients had low scores according to the NHB-ASCO-F and none could be graded as meaningful clinical beneficial according to the ESMO-MCBS. Despite the low incidence of NET, the heterogeneous patient popula-tion and the relatively long natural course, future studies on new treatment modalities should aim for high clinical benefit outcomes.

2

27


References 1. Mocellin S, Nitti D. Gastrointestinal carcinoid: epidemiological and survival evidence from a

large population-based study (n = 25 531). Ann Oncol 2013;24:3040-4. 2. Yao JC, Hassan M, Phan A, et al. One hundred years after “carcinoid”: epidemiology of and

prognostic factors for neuroendocrine tumors in 35,825 cases in the United States. J Clin Oncol 2008;26:3063-72.

3. Caplin ME, Pavel M, Cwikla JB, et al. Lanreotide in metastatic enteropancreatic neuroen-docrine tumors. N Engl J Med 2014;371:224-33.

4. Meyer T, Qian W, Caplin ME, et al. Capecitabine and streptozocin +/- cisplatin in advanced gastroenteropancreatic neuroendocrine tumours. Eur J Cancer 2014;50:902-11.

5. Pavel M, Oberg KE, Hainsworth JD, et al. Everolimus plus octreotide long-acting release (LAR) for the treatment of advanced neuroendocrine tumors (NET) associated with carci-noid syndrome (RADIANT-2): Updated overall survival results. Eur J Cancer 2013;49:S577.

6. Raymond E, Dahan L, Raoul JL, et al. Sunitinib malate for the treatment of pancreatic neuroendocrine tumors. N Engl J Med 2011;364:501-13.

7. Rinke A, Wittenberg M, Schade-Brittinger C, et al. Placebo-controlled, double-blind, pro-spective, randomized study on the effect of octreotide LAR in the control of tumor growth in patients with metastatic neuroendocrine midgut tumors (PROMID): Results of long-term survival. Neuroendocrinology 2017;104:26-32.

8. Strosberg J, El-Haddad G, Wolin E, et al. Phase 3 trial of 177Lu-dotatate for midgut neuro-endocrine tumors. N Engl J Med 2017;376:125-35.

9. Yao JC, Pavel M, Lombard-Bohas C, et al. Everolimus for the Treatment of Advanced Pancreatic Neuroendocrine Tumors: Overall Survival and Circulating Biomarkers From the Randomized, Phase III RADIANT-3 Study. J Clin Oncol 2016;34:3906-13.

10. Yao JC, Fazio N, Singh S, et al. Everolimus for the treatment of advanced, non-functional neuroendocrine tumours of the lung or gastrointestinal tract (RADIANT-4): a randomised, placebo-controlled, phase 3 study. Lancet 2016;387:968-77.

11. Pavel ME, Singh S, Strosberg JR, et al. Health-related quality of life for everolimus versus placebo in patients with advanced, non-functional, well-differentiated gastrointestinal or lung neuroendocrine tumours (RADIANT-4): a multicentre, randomised, double-blind, placebo-controlled, phase 3 trial. Lancet Oncol 2017;18:1411-22.

12. Yao JC, Guthrie KA, Moran C, et al. Phase III Prospective Randomized Comparison Trial of Depot Octreotide Plus Interferon Alfa-2b Versus Depot Octreotide Plus Bevacizumab in Patients With Advanced Carcinoid Tumors: SWOG S0518. J Clin Oncol 2017;35:1695-703.

13. Kulke MH, Siu LL, Tepper JE, et al. Future directions in the treatment of neuroendocrine tumors: consensus report of the National Cancer Institute Neuroendocrine Tumor clinical trials planning meeting. J Clin Oncol 2011;29:934-43.

14. Cherny NI, Sullivan R, Dafni U, et al. A standardised, generic, validated approach to stratify the magnitude of clinical benefit that can be anticipated from anti-cancer therapies: the Eu-ropean Society for Medical Oncology Magnitude of Clinical Benefit Scale (ESMO-MCBS). Ann Oncol 2015;26:1547-73.

15. Schnipper LE, Davidson NE, Wollins DS, et al. American Society of Clinical Oncology Statement: A conceptual framework to assess the value of cancer treatment options. J Clin Oncol 2015;33:2563-77.

Chapter 2

28

16. Schnipper LE, Davidson NE, Wollins DS, et al. Updating the American Society of Clinical Oncology Value Framework: Revisions and reflections in response to comments received. J Clin Oncol 2016;34:2925-34.

17. Kiesewetter B, Raderer M, Steger GG, et al. The European Society for Medical Oncology Magnitude of Clinical Benefit Scale in daily practice: a single institution, real-life experience at the Medical University of Vienna. ESMO Open 2016;1:e000066.

18. Wilson L, Lin T, Wang L, et al. Evaluation of the ASCO Value Framework for Anticancer Drugs at an Academic Medical Center. J Manag Care Spec Pharm 2017;23:163-9.

19. Ramage J, Punia P, Faluyi O, et al. UK phase IV, observational study to assess quality of life in patients (pts) with pancreatic neuroendocrine tumours (pNETS) receiving treatment with everolimus: The “real-world” OBLIQUE study. Ann Oncol 2016;27.

20. Booth CM, Eisenhauer EA. Progression-free survival: meaningful or simply measurable? J Clin Oncol 2012;30:1030-3.

21. Moodie EE, Karran JC, Shortreed SM. A case study of SMART attributes: a qualitative as-sessment of generalizability, retention rate, and trial quality. Trials 2016;17:242,016-1368-3.

22. van den Bent MJ, Brandes AA, Taphoorn MJ, et al. Adjuvant procarbazine, lomustine, and vincristine chemotherapy in newly diagnosed anaplastic oligodendroglioma: long-term follow-up of EORTC brain tumor group study 26951. J Clin Oncol 2013;31:344-50.

23. Bentley TGK, Cohen JT, Elkin EB, et al. Measuring the value of new drugs: Validity and reliability of 4 value assessment frameworks in the oncology setting. J Manag Care Spec Pharm 2017;23:S34-48.

24. Cheng S, McDonald EJ, Cheung MC, et al. Do the American Society of Clinical Oncol-ogy Value Framework and the European Society of Medical Oncology Magnitude of Clinical Benefit Scale Measure the Same Construct of Clinical Benefit? J Clin Oncol 2017;:JCO2016716894.

25. NCCN Evidence Blocks. National Comprehensive Cancer Network. (at https://www.nccn.org/evidenceblocks/).

26. Drug Abacus. Memorial Sloan Kettering Cancer Center. (2016, at http://www.drugabacus.org/).

27. The ICER approach. Institute for Clinicial and Economic Review. (2017, at https://icer-review.org/methodology/icers-methods/).

28. Schnipper LE, Schilsky RL. Converging on the Value of Value Frameworks. J Clin Oncol 2017;35:2732-2734.

2

29


Supplementary Table S1: Eligibility criteria for assessment with ESMO-MCBS and NHB of the revised ASCO framework

ESMO-MCBS NHB part of revised ASCO framework

Design Comparative outcome studies: (one of these)• Randomized design • Comparative cohort-design • Systematic review

New treatment compared with a prevailing standard of careA in a well-designed, well-conducted prospective randomized trialB.

Outcome Relative benefit of treatments; (one of these, if given as primary endpoint)• Survival • QoL• Surrogate outcomes for survival or QoL,• Treatment toxicities

Benefit: (one of these)• If available -> OS• If OS data is not available ->PFS• If OS and PFS data are not

available ->RR

Toxicity Toxicity data (Can up- or down grade the final score)

Toxicity data graded in accordance with WHO criteria in grade 1,2 and grade 3,4. (Can up- or down grade the final score)

Disease Solid cancers Advanced disease or potentially curable clinical presentationsB in medical oncology settingA.

Same clinical question evaluated in more studies

Results derived from well-powered registration trial should be given priority.

Only the highest quality evidence availableA.

Pre-planned subgroups

These can be scored with a maximum of three subgroups, scoring should be done separately when for more than one subgroup statistically significant results are reported.

NA

Post-hoc subgroup analysis

These are not graded, except for studies that incorporate collection of tissue samples to enable re-stratification based on new genetic or other biomarkers.

NA

In addition (Only if data are available)

QoL (Can up- or down grade the final score)

Bonus: • Tail of the curve• Palliation • QoL• Treatment free intervalB (Can upgrade the final score)

ASCO= American Society of Clinical Oncology, ESMO-MCBS= European Society of Medical Oncology- Magnitude of Clinical Benefit Scale, NA= not applicable, NHB= net health benefit, OS= overall survival, PFS=progression free survival, QoL=quality of life, RR=response rateA) In conceptual frameworkB) In revised framework

Chapter 2

30

2

31


Supp

lem

enta

ry T

able

S2:

Clin

ical

ben

efit o

f sys

tem

ic tr

eatm

ent f

or G

EP-N

ET a

ccor

ding

to E

SMO

-MC

BS

(ESM

O-M

CB

S) a

nd a

ccor

ding

ASC

O re

vise

d fr

amew

ork

(NH

B)

Inte

rven

tion

vs c

ontr

olES

MO

-MC

BS

ASC

O re

vise

d fr

amew

ork

Setti

ngPr

imar

y O

utco

me

HR

of

prim

ary

outc

ome/

RR

Into

xica

tion

data

Impr

ovem

ent

QO

L (Y

es/N

o)ES

MO

- M

CB

SC

linic

al

bene

fit

scor

e

Into

xica

tion

data

Bon

us

poin

tsN

HB

Ref

eren

ces

Ever

olim

us v

ersu

s pl

aceb

oPr

evio

us S

SA u

se

befo

re p

artic

ipat

ing

in

RAD

IAN

T_2

trial

PFS

NS

NS

NA

NA

0N

C0

NA

Anth

ony

(1)

Ever

olim

us v

ersu

s pl

aceb

oSS

A-na

ive

befo

re

parti

cipa

ting

in

RAD

IAN

T_2

trial

PFS

NS

NS

NA

NA

0N

C0

NA

Anth

ony

(1)

Oct

reot

ide

vs n

o or

ot

her t

reat

men

tIn

test

inal

neu

ro-

endo

crin

e tu

mor

PFS

NS

NA

NA

NA

0N

A0

NA

Arau

yo (2

)

Oct

reot

ide

and

inte

rfer

on v

s oc

treo

tide

Prog

ress

ive.

m

etas

tatic

GEP

-NET

OS

(and

TTF

)N

SN

SN

oN

A0

NC

0N

AAr

nold

(3)

Lanr

eotid

e au

toge

l vs

lanr

eotid

e m

icro

part

icel

s

Spor

adic

NET

. ne

w d

iagn

osis

or

prog

ress

ion

RR

NS

NS

NA

NA

0N

C0

NA

Baje

tta (4

)

Stre

ptoz

ocin

and

5FU

vs

inte

rfer

onU

nres

ecta

ble

mal

igna

nt c

arci

noid

tu

mor

PFS

NS

NS

No

NA

0N

C0

NA

Dah

an (5

)

Stre

ptoz

ocin

and

5FU

vs

dox

orub

icin

Unr

esec

tabl

e.

prog

ress

ive

met

asta

tic

carc

inoi

d.

OS

and

resp

onse

ra

te

NS

NS

NA

NA

0N

C0

NA

Engs

trom

(6)

Inte

rfer

on v

s la

nreo

tide

Prog

ress

ive.

m

etas

tatic

. tre

atm

ent-

naiv

e G

EP N

ET

RR

NS

NS

NA

NA

0N

C0

NA

Fais

s (7

)

Lanr

eotid

e vs

in

terf

eron

and

la

nreo

tide

Prog

ress

ive.

m

etas

tatic

. tre

atm

ent-

naiv

e G

EP N

ET

RR

NS

NS

NA

NA

0N

C10

NA

Fais

s (7

)

Oct

reot

ide

vs p

lace

boG

astro

-inte

stin

al N

ETN

umbe

r of

sym

ptom

epis

odes

an

d 5-

HIA

A

NA

NA

Yes.

in 2

/5

dom

ains

of

PAI

S

30

NA

10N

AJa

cobs

en (8

)

Chapter 2

30

2

31


Supp

lem

enta

ry T

able

S2:

Clin

ical

ben

efit o

f sys

tem

ic tr

eatm

ent f

or G

EP-N

ET a

ccor

ding

to E

SMO

-MC

BS

(ESM

O-M

CB

S) a

nd a

ccor

ding

ASC

O re

vise

d fr

amew

ork

(NH

B)

(con

tinue

d)

Inte

rven

tion

vs c

ontr

olES

MO

-MC

BS

ASC

O re

vise

d fr

amew

ork

Setti

ngPr

imar

y O

utco

me

HR

of

prim

ary

outc

ome/

RR

Into

xica

tion

data

Impr

ovem

ent

QO

L (Y

es/N

o)ES

MO

- M

CB

SC

linic

al

bene

fit

scor

e

Into

xica

tion

data

Bon

us

poin

tsN

HB

Ref

eren

ces

Inte

rfer

on a

lfa.

stre

ptoz

ocin

and

do

xoru

bici

n vs

in

terf

eron

Car

cino

id tu

mor

RR

NS

NS

NA

NA

0N

C0

NA

Jans

on (9

)

Oct

reot

ide

and

inte

rfer

on v

s oc

treo

tide

Car

cino

id tu

mor

of

mid

gut w

ith li

ver

met

asta

ses

Ris

k of

tu

mor

pr

ogre

ssio

n

0.28

(0

.16-

0.45

)N

SN

AN

AN

CN

A0

NA

Kölb

y (1

0)B

Telo

tris

tat E

thyl

and

SS

A vs

pla

cebo

and

SS

A

Met

asta

tic N

ET w

ith

carc

inoi

d sy

ndro

me

and

≥ 4

BM a

day

BM fr

eq

redu

ctio

n 44

vs

20%

NS

NA

NA

NA

NA

10N

AKu

lke

(11)

Telo

tris

tat E

thyl

and

SS

A vs

pla

cebo

and

SS

A

Met

asta

tic N

ET w

ith

carc

inoi

d sy

ndro

me

and

≥ 4

BM a

day

BM fr

eq

redu

ctio

n42

vs

20%

NS

NA

NA

NA

NA

10N

AKu

lke

(11)

Che

mot

hera

py b

efor

e ev

erol

imus

or p

lace

boC

hem

o-na

ive

parti

cipa

nts

of th

e R

ADIA

NT-

3 tri

al

PFS

0.32

(0

.21-

0.48

)N

AN

A3

44-2

00

24Lo

mba

rd-

Boha

s (1

2)

Che

mot

hera

py b

efor

e ev

erol

imus

or p

lace

boPa

rtici

pant

s w

ith p

rior

chem

othe

rapy

of t

he

RAD

IAN

T-3

trial

PFS

0.45

(0

.29-

0.70

)N

AN

A3

54.4

-20

1650

.4Lo

mba

rd-

Boha

s (1

2)

Stre

ptoz

ocin

and

C

yclo

phos

pham

ide

vs

Stre

ptoz

ocin

and

5FU

Unr

esec

atbl

e m

etas

tast

ic c

arci

noid

tu

mor

RR

26%

vs

33 %

No

NA

NA

00

00

Moe

rtel

’79

(13)

Stre

ptoz

ocin

and

5FU

vs

str

epto

zoci

nAd

vanc

ed is

let c

ell

carc

inom

aR

R63

vs

36%

No

NA

244

NC

0N

AM

oerte

l ’8

0 (1

4)C

hlor

ozot

ocin

vs

stre

ptoz

ocin

and

FU

vs

Str

epto

zoci

n an

d do

xoru

bici

n

Unr

esec

tabl

e or

m

etas

tatic

isle

t cel

l ca

rcin

oma.

RR

. but

als

o O

S is

giv

en1.

4yr v

s 2.

2 yr

St

rep/

FU v

s St

rep/

dox

NS

NA

415

1.9

NC

16N

AM

oerte

l ’92

(1

5)C

Chapter 2

32

2

33


Supp

lem

enta

ry T

able

S2:

Clin

ical

ben

efit o

f sys

tem

ic tr

eatm

ent f

or G

EP-N

ET a

ccor

ding

to E

SMO

-MC

BS

(ESM

O-M

CB

S) a

nd a

ccor

ding

ASC

O re

vise

d fr

amew

ork

(NH

B)

(con

tinue

d)

Inte

rven

tion

vs c

ontr

olES

MO

-MC

BS

ASC

O re

vise

d fr

amew

ork

Setti

ngPr

imar

y O

utco

me

HR

of

prim

ary

outc

ome/

RR

Into

xica

tion

data

Impr

ovem

ent

QO

L (Y

es/N

o)ES

MO

- M

CB

SC

linic

al

bene

fit

scor

e

Into

xica

tion

data

Bon

us

poin

tsN

HB

Ref

eren

ces

Inte

rfer

on v

s St

rept

ozoc

in a

nd 5

FUC

arci

noid

tum

orR

R o

r sta

ble

dise

ase

0 vs

20

%N

SN

A2

14N

C0

NA

Obe

rg ’8

9 (1

6)St

rept

ozoc

in a

nd

5FU

vs

stre

ptoz

ocin

an

d 5F

U fo

llow

ed b

y in

terf

eron

Car

cino

id tu

mor

with

liv

er m

etas

tase

s.R

R. b

ut a

lso

OS

is g

iven

8 vs

72 m

p<

0.00

1N

SN

A4

NC

NC

20N

AO

berg

‘91

(17)

D

Stre

ptoz

ocin

and

5FU

vs

firs

t-lin

e in

terf

eron

Car

cino

id tu

mor

with

liv

er m

etas

tase

s.R

R. b

ut a

lso

OS

is g

iven

8 vs

56

mN

SN

A4

NC

NA

20N

AO

berg

‘91

(17)

D

Lanr

eotid

e vs

oc

treo

tide

Car

cino

id tu

mor

with

ca

rcin

oid

synd

rom

e.Q

oLdi

sapp

eara

nce

of fl

ushe

s

NS

NS

No

NA

0N

A0

NA

O’ T

oole

(1

8)

Oct

reot

ide

LAR

vs

no

or o

ther

trea

tmen

tM

etas

tasi

zed

NET

pa

tient

s of

≥ 6

5 ye

ars

regi

ster

ed in

the

SEER

dat

abas

e

OS

0.61

(0.4

7-0.

79)

NA

NA

439

NA

0N

ASh

en

(19)

Oct

reot

ide

LAR

vs

no tr

eatm

ent w

ith

octr

eotid

e LA

R

NET

pat

ient

s ≥

65

year

regi

ster

ed in

the

SEER

dat

abas

e. w

ith

loco

-regi

onal

dis

ease

.

OS

NS

NA

NA

NA

0N

A0

NA

Shen

(19)

Oct

reot

ide

LAR

w

ith p

revi

ous

SSA

vs O

ctre

otid

e LA

R

with

out p

revi

ous

SSA

Fore

gut.

mid

gut o

r hi

ndgu

t NET

-pat

ient

s in

clud

ed in

the

plac

ebo

arm

of t

he

RAD

IAN

T-2

trial

.

OS

33.5

m v

s 50

.6 m

NA

NA

NA

NA

NA

NA

NA

Stro

sber

g (2

0)E

Stre

ptoz

ocin

/5FU

vs

dox

orub

icin

/5FU

fo

llow

ed b

y D

TIC

if P

D

Unr

esec

tabl

e.

adva

nced

car

cino

id

tum

ors

RR

or O

S15

.7m

vs

24.3

m O

S p=

0.02

67

NS

NA

454

.8N

C0

NA

Sun

(21)

Chapter 2

32

2

33


Supp

lem

enta

ry T

able

S2:

Clin

ical

ben

efit o

f sys

tem

ic tr

eatm

ent f

or G

EP-N

ET a

ccor

ding

to E

SMO

-MC

BS

(ESM

O-M

CB

S) a

nd a

ccor

ding

ASC

O re

vise

d fr

amew

ork

(NH

B)

(con

tinue

d)

Inte

rven

tion

vs c

ontr

olES

MO

-MC

BS

ASC

O re

vise

d fr

amew

ork

Setti

ngPr

imar

y O

utco

me

HR

of

prim

ary

outc

ome/

RR

Into

xica

tion

data

Impr

ovem

ent

QO

L (Y

es/N

o)ES

MO

- M

CB

SC

linic

al

bene

fit

scor

e

Into

xica

tion

data

Bon

us

poin

tsN

HB

Ref

eren

ces

Y-D

OTA

TOC

vs

Y-D

OTA

TOC

and

LU

-D

OTA

TOC

Prog

ress

ive.

m

etas

tatic

ne

uroe

ndoc

rine

tum

or.

OS

0.64

(0.4

7-0.

88)

NS

NA

436

NC

0N

AVi

llard

(22)

Lanr

eotid

e vs

pla

cebo

NET

pat

ient

s w

ith

carc

inoi

d sy

ndro

me

Sym

ptom

co

ntro

l(d

ays

of

octre

otid

e)

NS

NS

NS

NA

NA

NA

0N

AVi

nik

(23)

Pasi

reot

ide

LAR

vs

octr

eotid

e LA

RC

arci

noid

tum

or

of d

iges

tive

tract

w

ith in

adeq

uate

ly

cont

rolle

d di

arrh

ea

and/

or fl

ushi

ng.

Sym

ptom

co

ntro

lN

SN

SN

A3

43.2

NC

0N

AW

olin

(24)

Oct

reot

ide

and

beva

cizu

mab

vs

PEG

-inte

rfero

n an

d O

ctre

otid

e

Met

asta

tic c

arci

noid

tu

mor

Res

pons

e (a

nd

PFS)

NS

NS

NA

NA

0N

C0

NA

Yao

’08

(25)

Ever

olim

us w

ith

octr

eotid

e LA

R v

s ev

erol

imus

with

out

octr

eotid

e LA

R

Adva

nced

pan

crea

tic

NET

with

pro

gres

sive

di

seas

e af

ter

chem

othe

rapy

RR

NA

NS

NA

NA

0N

C0

NA

Yao

’10

(26)

F

ASC

O=

Amer

ican

Soc

iety

of C

linic

al O

ncol

ogy.

BM

= bo

wel

mov

emen

ts. B

M fr

eq re

duct

ion=

repo

nse.

defi

ned

as a

bm

freq

uenc

y re

duct

ion

≥ 30

% fr

om b

asel

ine

for ≥

50%

of t

he

treat

men

t per

iod.

ESM

O-M

CBS

= Eu

rope

an S

ocie

ty o

f Med

ical

Onc

olog

y- M

agni

tude

of C

linic

al B

enefi

t Sca

le. G

EP-N

ET=g

astro

inte

stin

al a

nd/ o

r pan

crea

tic n

euro

-end

ocrin

e tu

mor

. G

I=ga

stro

-inte

stin

al. H

R=

haza

rd ra

tio. N

ET=n

euro

endo

crin

e tu

mor

. vs=

vers

us. L

AR=l

ong

actin

g re

agen

ts. N

A= n

ot a

pplic

able

NC

. no

cons

ensu

s. N

D=

no im

prov

emen

t or n

o im

-po

rtant

diff

eren

ce. N

S=no

sig

nific

ant d

iffer

ence

. OS=

over

all s

urvi

val.

PFS=

prog

ress

ion

free

surv

ival

. QoL

=qua

lity

of lif

e. R

R=r

espo

nse

rate

. SSA

=som

atos

tatin

ana

logu

e vs

= ve

rsus

.A

Firs

t aut

hor.

Ref

eren

ces

are

repo

rted

in th

e su

pple

men

tary

refe

renc

e lis

t.B

Alth

ough

the

risk

of tu

mor

pro

gres

sion

is s

igni

fican

t diff

eren

t. th

is e

ndpo

int c

an n

ot b

e us

ed in

the

ESM

O-M

CBS

. nor

in th

e AS

CO

Fra

mew

ork.

Som

e ce

nter

s us

ed th

is e

ndpo

int a

s be

ing

a H

R o

f PFS

. or u

sed

inst

ead

of th

is e

ndpo

int t

he re

spon

se ra

te (a

lthou

gh n

o p-

valu

e or

sig

nific

ance

leve

l was

giv

en w

ith th

e re

spon

se ra

te).

C A

lthou

gh th

e tim

e to

tum

or p

rogr

essi

on is

sig

nific

ant d

iffer

ent t

his

endp

oint

can

not

be

used

in th

e ES

MO

-MC

BS a

nd A

SCO

Fra

mew

ork.

In A

SCO

fram

ewor

k ov

eral

l sur

viva

l is

not p

erm

itted

bec

ause

cro

ss o

ver.

D P

atie

nts

wer

e no

t ran

dom

ized

. but

ass

igne

d to

a tr

eatm

ent.

This

pro

babl

y ha

ve le

d to

bia

s.

E Pa

tient

s w

ere

divi

ded

base

d on

the

treat

men

t the

par

ticip

ant h

ad b

efor

e in

clus

ion

in th

e R

ADIA

NT-

2 tri

al. N

o st

atis

tical

ana

lysi

s is

per

form

ed b

etw

een

the

two

grou

ps.

F Pa

tient

s w

ere

stra

tified

no

stat

istic

al a

naly

sis

is p

erfo

rmed

bet

wee

n th

e tw

o st

rata

Chapter 2

34

Supplementary Figure S1. Inclusion process for the literature analysis

1942 articles identified by database search:-1442 Embase -500 PubMed

1676 articles

266 Duplicates excluded

1453 ‘Clearly not relevant’ articles excluded based on title and abstract

39 articles remaining for analysis including 35 trials

223 articles

189 articles excluded based on:

The clinical question of 43 articles, was published later in a better article or with a better trial design. In 83 articles the main subject was not the comparison of intervention 8 Articles showed only a multivariate analysis 1 Article was not only about NET 20 Articles were post-hoc analyses, without comparison of intervention 21 Articles did not have a cohort design, a randomized trial design, or a design of a meta-analysis. 7 Articles only showed baseline results or the planned study 1 Article used an outcome that was not eligible for scoring (i.e. blood flow in liver metastases) 5 Only abstract available

Nog precies natellen

5 articles with publication date after literature search included later, 4 Phase III trials, 1 post-hoc analysis of predefined quality of life data

Supplementary Figure S1. Inclusion process for the literature analysis

2

35


Figure 2. Eligibility criteria for a valid evaluation with the tools

Reference

Both thools

Both tools

Both tools

ASCO

ASCO

ASCO

ESMO-M

CBS

ESMO-M

CBS

A B C D E F G H Eligibility criteria that shouldCaplin (1) be met for a valid score:Meyer (2) Eligibility criteria are Pavel (3,4) not met.Raymond (5) Eligibility criteria are Rinke (6,7) partly met.Strosberg (8) All eligibility criteria Yao’11, Yao‘16 (9,10) are met.Yao’16 (11)Yao’17 (12)Wolin (34)Faiss (19)Arnold (15)Dahan (17)Moertel ’92 (26) Jacobsen (20) Janson (21)Bajetta (16)Kölby (22)Moertel ’79 (24)Sun (32)Moertel ’80 (25)Oberg ’89 (27)Yao ’10 (35)O’ Toole (29)Shen (30)Villard (33)Lombard-Bohas (23)Arauyo (14)Anthony (13)Strosberg (31) Oberg ‘91 (28)Engstrom (18)

A; More than 50% of participants has another NET than a gastro-intestinal or pancreatic NET B; Sample size of pre-planned power analysis not reached C; Same clinical question evaluated in more studiesD; Toxicity data are not available and gradedE; Outcome data availableF; No prospective, randomized designG; No randomized, comparative cohort design, or meta-analysisH; No toxicity data available Figure 2. Eligibility criteria for a valid evaluation with the tools

Chapter 2

36

Supplementary data; Extensive methods

Search strategyTrials were searched in the databases PubMed and EMBASE between Aug 1st 2015 to Jan 31th 2016. The following search terms were used in various combinations ‘neuroen-docrine’, ‘tumor’, ‘pNET’, ‘neoplasm’, ‘survival’, ‘disease-free survival’, ‘quality of life’, ‘QOL’, ‘toxicity’, ‘adverse events’, ‘chemotherapy’, ‘cytostatic’, ‘interferon’, ‘somatostatin analogue’, ‘everolimus’, ‘sunitinib’, ‘indium’, ‘lutetium’, ‘peptide receptor radionuclide’, ‘controlled trial’ and ‘clinical trial’. The articles that were found were screened using title and abstract to select trials published in the English language and comparing systemic treatment modalities for the treatment of GEP-NET in humans. Furthermore the refer-ence lists of the European Neuroendocrine Tumor Society (ENETS), National Compre-hensive Cancer Network, (NCCN) and North American Neuroendocrine Tumor Society (NANETS), guidelines were reviewed for comparative trials that support the guidelines (suppl ref 27-29). Also reference lists of included trials and conference abstracts were reviewed for additional relevant articles.

Selection criteria for trialsComparative trials investigating systemic antitumor treatment for GEP-NET patients were analyzed. Criteria for assessment are summarized in Table S1. Trials were eligible for grading with the ESMO-MCBS if either a randomized or comparative cohort design was used or if a meta-analysis was used reporting a statistically significant benefit in any of the evaluated outcomes. With the NHB of the revised ASCO framework only randomized controlled trials (RCTs) could be graded. Furthermore for both tools only well-powered trials evaluating survival, QoL, surrogate outcomes of survival (disease-free interval, event-free survival, time to response, PFS and time to progression), or treatment toxicity could be graded. When more than one trial, that fulfilled the selec-tion criteria, analyzed the same clinical question, results derived from well powered registration trials, prevailed. Trials were selected if at least 50% of participants were diagnosed with GEP-NET, and the other participants had a NET of unknown or other origin. Abstracts reporting clinical trials in GEP-NET patients were only permitted for analysis if they showed additional data to an already published article.

ESMO-MCBS and the NHB of the revised ASCO framework ESMO-MCBS grades in the non-curative setting ranges from 1-5, with grade 4 and 5 representing meaningful clinical benefit. To evaluate different endpoints three forms are available; form 2a and 2b for ‘trials with primary endpoint OS’, and ‘PFS’, respectively and form 2c for trials with ‘with primary endpoint other than OS or PFS or equivalence studies’. The NHB of the revised ASCO framework consists of the clinical benefit score

2

37


(CBS), (i.e. the hazard ratio or median of OS or PFS, or response rate (RR)) toxicity and bonus points (for long term disease control, palliation, treatment free interval and QoL). The NHB of the revised ASCO framework ranges from -20 to 180, with a higher score representing a better NHB, no cutoff value was provided to define clinical benefit. All relevant comparative trials were assessed according to both tools. Relevant trials that did not meet all criteria as mentioned in Table 1 were assessed separately, based on the available data.

Assessment of trialsFour members from all Dutch European Neuroendocrine Tumor Society (ENETS) cen-ters of excellence (Academic Medical Center, Erasmus Medical Center, Netherlands Cancer Institute, and University Medical Center Groningen) independently scored the trials according to the ESMO-MCBS and the NHB of the revised ASCO framework. After obtaining scores of these assessors we noticed a wide variation in results. During a consensus meeting additional agreements were defined. Next, trials were assessed again according to the additional agreements. The number of trials to which the same score was awarded between three or four assessors was registered after each of the two scoring sessions. Two phase III trials were published after the consensus meeting and were assessed according our agreements. Therefore an additional assessment session was not necessary for these trials.

Chapter 2

38

Supplementary data; references

References 1. Anthony LB, Pavel ME, Hainsworth JD, et al. Impact of previous somatostatin analogue use

on the activity of everolimus in patients with advanced neuroendocrine tumors: Analysis from the phase III RADIANT-2 trial. Neuroendocrinology 2015;102:18-25.

2. Araujo PB, Cheng S, Mete O, et al. Evaluation of the WHO 2010 grading and AJCC/UICC staging systems in prognostic behavior of intestinal neuroendocrine tumors. PLoS One 2013;8:e61538.

3. Arnold R, Rinke A, Klose KJ, et al. Octreotide versus octreotide plus interferon-alpha in endocrine gastroenteropancreatic tumors: a randomized trial. Clin Gastroenterol Hepatol 2005;3:761-71.

4. Bajetta E, Procopio G, Catena L, et al. Lanreotide autogel every 6 weeks compared with Lanreotide microparticles every 3 weeks in patients with well differentiated neuroendocrine tumors: a Phase III Study. Cancer 2006;107:2474-81.

5. Dahan L, Bonnetain F, Rougier P, et al. Phase III trial of chemotherapy using 5-fluoro-uracil and streptozotocin compared with interferon alpha for advanced carcinoid tumors: FNCLCC-FFCD 9710. Endocr Relat Cancer 2009;16:1351-61.

6. Engstrom PF, Lavin PT, Moertel CG, Folsch E, Douglass HO,Jr. Streptozocin plus fluoroura-cil versus doxorubicin therapy for metastatic carcinoid tumor. J Clin Oncol 1984;2:1255-9.

7. Faiss S, Pape UF, Bohmig M, et al. Prospective, randomized, multicenter trial on the antiproliferative effect of lanreotide, interferon alfa, and their combination for therapy of metastatic neuroendocrine gastroenteropancreatic tumors--the International Lanreotide and Interferon Alfa Study Group. J Clin Oncol 2003;21:2689-96.

8. Jacobsen MB, Hanssen LE. Clinical effects of octreotide compared to placebo in patients with gastrointestinal neuroendocrine tumours. Report on a double-blind, randomized trial. J Intern Med 1995;237:269-75.

9. Janson ET, Ronnblom L, Ahlstrom H, et al. Treatment with alpha-interferon versus alpha-interferon in combination with streptozocin and doxorubicin in patients with malignant carcinoid tumors: a randomized trial. Ann Oncol 1992;3:635-8.

10. Kolby L, Persson G, Franzen S, Ahren B. Randomized clinical trial of the effect of inter-feron alpha on survival in patients with disseminated midgut carcinoid tumours. Br J Surg 2003;90:687-93.

11. Kulke MH, Horsch D, Caplin ME, et al. Telotristat ethyl, a tryptophan hydroxylase inhibitor for the treatment of carcinoid syndrome. J Clin Oncol 2017;35:14-23.

12. Lombard-Bohas C, Yao JC, Hobday T, et al. Impact of prior chemotherapy use on the efficacy of everolimus in patients with advanced pancreatic neuroendocrine tumors: a subgroup analysis of the phase III RADIANT-3 trial. Pancreas 2015;44:181-9.

13. Moertel CG, Hanley JA. Combination chemotherapy trials in metastatic carcinoid tumor and the malignant carcinoid syndrome. Cancer Clin Trials 1979;2:327-34.

14. Moertel CG, Hanley JA, Johnson LA. Streptozocin alone compared with streptozocin plus fluorouracil in the treatment of advanced islet-cell carcinoma. N Engl J Med 1980;303:1189-94.

15. Moertel CG, Lefkopoulo M, Lipsitz S, Hahn RG, Klaassen D. Streptozocin-doxorubicin, streptozocin-fluorouracil or chlorozotocin in the treatment of advanced islet-cell carcinoma. N Engl J Med 1992;326:519-23.

16. Oberg K, Norheim I, Alm G. Treatment of malignant carcinoid tumors: a randomized con-trolled study of streptozocin plus 5-FU and human leukocyte interferon. Eur J Cancer Clin Oncol 1989;25:1475-9.

17. Oberg K, Eriksson B. The role of interferons in the management of carcinoid tumours. Br J Haematol 1991;79 Suppl 1:74-7.

18. O’Toole D, Ducreux M, Bommelaer G, et al. Treatment of carcinoid syndrome: a prospective crossover evaluation of lanreotide versus octreotide in terms of efficacy, patient accept-ability, and tolerance. Cancer 2000;88:770-6.

19. Shen C, Shih YC, Xu Y, Yao JC. Octreotide long-acting repeatable use among elderly pa-tients with carcinoid syndrome and survival outcomes: a population-based analysis. Cancer 2014;120:2039-49.

20. Strosberg JR, Yao JC, Bajetta E, et al. Efficacy of octreotide long-acting repeatable in neuroendocrine tumors: RADIANT-2 placebo arm post hoc analysis. Endocr Relat Cancer 2015;22:933-40.

21. Sun W, Lipsitz S, Catalano P, Mailliard JA, Haller DG, Eastern Cooperative Oncology Group. Phase II/III study of doxorubicin with fluorouracil compared with streptozocin with fluorouracil or dacarbazine in the treatment of advanced carcinoid tumors: Eastern Coop-erative Oncology Group Study E1281. J Clin Oncol 2005;23:4897-904.

22. Villard L, Romer A, Marincek N, et al. Cohort study of somatostatin-based radiopeptide therapy with [90Y-DOTA]-TOC versus [90Y-DOTA]-TOC plus [177Lu-DOTA]-TOC in neuroen-docrine cancers. J Clin Oncol 2012;30:1100-6.

23. Vinik AI, Wolin EM, Liyanage N, Gomez-Panzani E, Fisher GA, ELECT Study Group *. Eval-uation of lanreotide depot/autogel efficacy and safety as a carcinoid syndrome treatment (elect): a randomized, double-blind, placebo-controlled trial. Endocr Pract 2016;22:1068-80.

24. Wolin EM, Jarzab B, Eriksson B, et al. Phase III study of pasireotide long-acting release in patients with metastatic neuroendocrine tumors and carcinoid symptoms refractory to available somatostatin analogues. Drug Des Devel Ther 2015;9:5075-86.

25. Yao JC, Phan A, Hoff PM, et al. Targeting vascular endothelial growth factor in advanced carcinoid tumor: a random assignment phase II study of depot octreotide with bevacizumab and pegylated interferon alpha-2b. J Clin Oncol 2008;26:1316-23.

26. Yao JC, Lombard-Bohas C, Baudin E, et al. Daily oral everolimus activity in patients with metastatic pancreatic neuroendocrine tumors after failure of cytotoxic chemotherapy: a phase II trial. J Clin Oncol 2010;28:69-76.

27 ENETS Guidelines 2016. (at http://www.enets.org/current_guidelines.html). 28. NCCN Guidelines 2017. (at https://www.nccn.org/professionals/physician_gls/pdf/neuroen-

docrine.pdf). 29. Kunz PL, Reidy-Lagunes D, Anthony LB, et al. Consensus guidelines for the management

and treatment of neuroendocrine tumors. Pancreas 2013;42:557-77

3a

Web-based information and support for patients with a newly diagnosed

neuroendocrine tumor: a feasibility study

G. Bouma1, L.D. de Hosson1, C.E. van Woerkom1, H. van Essen1, G.H. de Bock2, J.M. Admiraal1, A.K.L. Reyners1, A.M.E. Walenkamp1

1Department of Medical Oncology, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands

2Department of Epidemiology, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands

Chapter 3a

42

Abstract

Purpose Patients with a neuroendocrine tumor (NET) frequently experience physical and psy-chosocial complaints. Novel strategies to provide information to optimize supportive care in these patients are of interest. The aim of this study was to examine whether the use of a web-based system consisting of self-screening of problems and care needs, patient education and self-referral to professional health care is feasible in patients with NET and to evaluate their opinion on this.

Methods Newly diagnosed patients with NET were randomized between standard care (n=10) or intervention with additional access to the web-based system (n=10) during 12 weeks. Patients completed questionnaires regarding received information, distress, quality of life (QoL) and empowerment. The intervention group completed a semi-structured interview to assess patients’ opinion on the web-based system.

Results The participation rate was 77% (20/26 invited patients) with no dropouts. The use of the web-based system had a negative effect on patients’ perception and satisfaction of received information (range Cohen’s d -0.88 to 0.13). Positive effects were found for distress (Cohen’s d 0.75), global QoL (subscale European Organization for Research and Treatment of Cancer (EORTC) QLQ-C30, Cohen’s d 0.46), resolving problems with social functioning and finding information (subscales EORTC QLQ-GINET 21, Cohen’s d 0.69 respectively 1.04), and feeling informed (subscale empowerment questionnaire, Cohen’s d 0.51). The interview indicated that the web-based system was of additional value to standard care.

Conclusions Use of this web-based system is feasible. Contradictory effects on informing and sup-porting patients with NET were found and should be subject of further research.

Keywords: neuroendocrine tumor, information, support, web-based, internet.

3

43

Web-based information and support for patients with a newly diagnosed neuroendocrine tumor

Introduction

Neuroendocrine tumors (NETs) are rare tumors arising from secretory cells of the neu-roendocrine system. Secretion of hormones and/or amines can result in a broad set of symptoms (1,2). Patients can experience various symptoms as a consequence of the presence of the tumor as well as the release of the hormones secreted by the tumor. Also, side effects of the treatment can bother them. Moreover, diagnosis and treatment of cancer often results in physical, psychological, practical, social and spiritual concerns (3). Studies in patients with NETs demonstrate that these concerns lead to lower health related quality of life (QoL) compared with the general population (4,5). In general, when patients with cancer are properly informed, they experience a better health related QoL and less depression and anxiety (6,7). In addition, providing information also results in an increased participation of patients in decision making, greater satisfaction with care, lower levels of distress and improved sense of control (7-10).Among patients with cancer, the increasing role of internet as a source of cancer related information is recognized. Web-based patient education and support have become widely accepted (11-13). We developed a web-based system for NET patients with the aim to examine its feasibility and to evaluate patient’s opinion on this. Secondary aims were to explore the effects on patients’ perception and satisfaction of received information, distress, health related QoL, and empowerment. The web-based system allows patients to self-screen for physical and psychosocial problems, to get tailored patient education on reported problems and if necessary, to refer themselves to care.

Methods

ParticipantsTo be eligible for this study, patients had to be newly diagnosed with a NET grade 1 or 2 (according to the World Health Organization 2010 classification) within 3 months before study participation. Any primary site or disease stage was allowed. Patients had to be under surveillance or treatment at the Department of Medical Oncology in the University Medical Center Groningen (UMCG). Additional inclusion criteria were: 18 years of age or older and ability to read and write Dutch. Not eligible were patients with an estimated life expectancy of less than 3 months or a second primary tumor with active follow-up or treatment. All eligible patients were invited to participate. Recruitment occurred at the outpatient clinic during October 2013 to January 2014. The study was approved by the medical ethical committee of the UMCG and registered in ClinicalTrials.gov (NCT01849523). All patients gave written informed consent.

Chapter 3a

44

Study procedureFor detailed description of study procedures see Supplemental material. Eligible pa-tients were informed about the study by their treating medical oncologist during a follow-up visit or by phone if a follow-up visit was not planned at short notice. Patients were asked if they could be approached for study participation by one of the investigators. If this was allowed, the investigator informed the patient orally after the follow-up visit or by phone. Patients were informed about the aims, study procedures and feasibility design of this study. Following obtained informed consent, patients were randomized in a 1:1 ratio to the standard care group or the intervention group that received standard care with additional access to the web-based system. At baseline, all patients received questionnaires by mail regarding socio-demographic characteristics, health care/inter-net use, perception and satisfaction of received information (EORTC QLQ-INFO25) and QoL (EORTC QLQ-C30, EORTC QLQ-GINET21). At baseline, the standard care group received the paper-and-pencil version of the Distress Thermometer (DT) and Problem List (PL) by mail. The intervention group had to complete the online version of the DT and PL (content identical to the paper-and-pencil-version). When questionnaires were completed, the intervention group received instructions how to use the web-based system together with log-in information including a personal username and password. At week 12, all patients received questionnaires by mail regarding frequency of visits to psychosocial and/or allied health care professionals, perception and satisfaction of received information, QoL and empowerment (Constructs Empowering Outcomes, (CEO) questionnaire). The standard care group also received the paper-and-pencil version of the DT and PL by mail. The intervention group was reminded to complete the online version of the DT and PL after 12 weeks. At end of the study, the patients assigned to the intervention group were also invited for a semi-structured interview with the investigator regarding their opinion on the web-based system. Randomization outcome was not blinded and also notified to the treating medical oncolo-gist. In this manner, all patients could discuss their questions about the retrieved informa-tion (e.g. from the web-based system, information leaflets) with their treating medical on-cologist, oncology nurse, other health care professionals and/or allied health care workers.

Standard care and study interventionStandard care for newly diagnosed patients with a NET during the first visit(s) at the Department of Medical Oncology includes verbal information about the diagnosis, evaluating complaints/problems and a physical examination by their treating medical oncologist. Also, newly diagnosed patients visit an oncology nurse with experience in care for patients with NET. All patients receive at diagnosis information leaflets of the Dutch Patient Neuroendocrine Tumor Foundation (14). During follow-up visits, the medi-cal oncologist evaluates and discusses the general well-being, test results, treatment

3

45


(e.g. options, side effects), answers questions of the patient and performs a physical examination. If physical and/or psychosocial problems require more in-depth discus-sion, investigation or treatment, patients can receive a consultation with the oncology nurse or other health care professionals. In the intervention group, patients received access to the web-based information and sup-port system in addition to standard care during the 12-week study period. The web-based system incorporates elements of self-screening for problems and care needs, tailored support in self-help, patient education and communication with a health care professional. Patients screened themselves for psychosocial/physical problems and care needs by filling out the DT and PL. Immediately after completion of the DT, automated personal-ized feedback was received upon their reported distress score. For example, feedback emphasizes the possibility for a referral to a health care professional for moderate or severe distress. Thereafter, patients received tailored feedback on physical/psychosocial problems they had reported on the PL. This feedback comprises background informa-tion about the problem, advice on how to cope with the problem (‘self-help advice’) and which health care professional can be consulted when self-help insufficiently alleviates problems. Online completed DT and PL questionnaires with feedback were online saved and coded to ensure that patients could access the feedback and information later.Also, general information about NETs and diagnostic procedures, treatments, and their side effects is provided by the web-based system. Patients could send an e-mail or request for a telephone call to communicate with the investigator (experienced physi-cian in treating patients with NET) in case of questions, problems or a referral wish. A referral wish to a professional or allied health care worker was arranged by the treating medical oncologist.

MeasuresPatient characteristics including socio-demographic characteristics, use of previous and current psychosocial and/or allied health care, acquaintence with internet and use of internet to search cancer-related information were assessed by a self-report question-naire. Information on disease-related characteristics was collected from medical records.Participation and dropout rates and reasons for declining study participation and drop-out were tracked. Patients’ perceived amount and satisfaction of received information was measured with the Dutch EORTC QLQ-INFO25 questionnaire, which is a validated questionnaire to evaluate the information received by cancer patients (15). This questionnaire comprises multi-items grouped into four information provision subscales ‘perceived receipt of in-formation about the disease’, ‘medical tests’, ‘treatment’ and ‘other care services’ and single-items on ‘having received information’ in different manners (e.g. written, online, on CD) and on ’satisfaction with, amount of and helpfulness of information’. In addition,

Chapter 3a

46

two open questions investigate topics of which patients would like to receive more or less information. All responses are ranged on a four-point Likert scale, except for four single items that have a ‘yes/no’-scale. Distress was measured using the validated Dutch DT and PL (3). The DT consists of a single item asking patients to indicate the amount of overall distress experienced during the past week on an 11-point scale (range ‘0’ no distress – ‘10’ extreme distress). The PL identifies existing practical, family/social, emotional, religious/spiritual and physi-cal problems and was adapted by the investigators according to the issues which are prevalent among and relevant to patients with NETs (in total 31 items). Patients could also add a problem that was not mentioned in the PL. For each reported problem, pa-tients rated the associated experienced amount of distress (range ‘0’ no distress – ‘10’ extreme distress). The PL ends with the question whether the patient has a referral wish to an oncology nurse, peer NET sufferer group, a psychosocial (psychologist, social or pastoral worker) or allied health care professional (e.g. physical therapist, dietician).QoL was measured with the validated Dutch EORTC QLQ-C30 (version 3) (16) and Dutch EORTC QLQ-GINET21 (17). The EORTC QLQ-C30 measures cancer-specific QoL comprising five functional scales (physical, role, cognitive, emotional, and social), three symptom scales (fatigue, pain and nausea/vomiting), six single symptoms items (dyspnea, insomnia, loss of appetite, constipation, diarrhea and financial impact) and a global health status/QoL scale. Responses are given on a 4-point Likert scale (range ‘1’ not at all - ‘4’ very much) with higher scores indicating higher QoL at the functional scales and higher symptom burden for the symptom scales. The global health status/QoL scale is scored on a 7-point scale (range ‘1’ very poor - ‘7’ excellent) with higher scores indicating higher QoL. The EORTC QLQ-GINET21 is a disease-specific QoL questionnaire for patients with a gastro-intestinal NET to supplement the EORTC QLQ-C30 questionnaire and assesses specific disease symptoms, side effects of treatment, body image, disease related worries, social functioning, communication and sexuality. Responses are given on a 4-point Likert scale (range ‘1’ not at all - ‘4’ very much) with higher scores indicating higher symptom burden for the symptom scales.Empowerment was measured by the Dutch CEO questionnaire, which is developed and tested in different online support groups including breast cancer patients (11,18). In the questionnaire, ‘online support groups’ was modified in ‘website’ for the intervention group and in ‘standard care’ for the control group. The domains ‘being better informed’, ‘feeling more confident about the relationship with their physician’ ‘improved acceptance of the illness’, ‘feeling more confident about the treatment’ and ‘increased optimism and control over the future’ were measured since these domains were expected to be influenced by the web-based system. Patients’ opinion regarding satisfaction with the web-based system was assessed with a semi-structured interview by phone with participants in the intervention group. Ten

3

47


open-ended, both qualitative and quantitative questions were used and when relevant the interviewer probed the participant with follow-up questions to include topics not es-tablished beforehand. These questions comprised the following issues: contents, lay-out, user-friendliness, perceived usability, usage and future recommendations. The interviews were recorded and afterwards all interviews were transcribed in English. All topics per individual interview related to the purpose of the research were noted and categorized.

Statistical analysisThe purpose of this study was to explore the feasibility and to evaluate patient’s opinion on the web-based system on above mentioned endpoints. Twenty patients with NET were recruited and randomized. Descriptive statistics (e.g. means, ranges, frequencies) were calculated for all measures. Missing values for the EORTC questionnaires were handled according to the EORTC guidelines (19). Raw scores of all scales and items of the EORTC questionnaires were linearly transformed to a score ranging from 0 to 100 (19). Scores of the different problem/empowerment domains of the PL/CEO question-naire were calculated by taking the mean of the total scores of the items in each domain. Effect sizes for change scores on each endpoint were determined (Cohen’s d) by using the mean difference between post- and pre-intervention scores of both groups and dividing this difference by the pooled standard deviation. Given the retrospective design of the CEO questionnaire with only a post-intervention measurement, effect sizes of the outcomes of the CEO were calculated by using the mean difference between post-intervention scores of both groups and dividing this difference by the accompanying pooled standard deviation. All effect sizes are reported so that a positive effect size indicates a desired direction (i.e. improvement of outcome in favor of the intervention group). An effect size (ES) be-tween 0.20-0.50 is defined as small, between 0.50-0.80 as medium and ≥0.80 as large. Analyses were performed using the software package SPSS, version 22 for Windows (SPSS, Inc, Chicago, IL, USA).

Results

ParticipantsTwenty of the 26 invited eligible patients gave informed consent and all patients com-pleted the full study period. Reasons for declining study participation were no access to a computer and internet (n=2), no interest in information and filling out questionnaires (n=1), feeling overburdened (n=2) and not willing to be further confronted with the disease (n=1).

Chapter 3a

48

Patient and tumor characteristics are summarized in Table 1. There was a minor imbal-ance between the intervention and control group regarding age, educational level and primary tumor localization. The intervention group consisted of younger, higher edu-cated patients and more patients with a NET of the lung. During the study, six patients (n=1 in the standard care group) started treatment with a somatostatin analogue and two patients (both in standard care group) with everolimus. Three patients (n=1 in the standard care group) underwent curative surgery and two patients in the intervention group underwent palliative surgery during the study period.

Table 1. Patient and tumor characteristicsStandard care group

(n=10)Na

Intervention group (n=10)

Na

Median age in years (range) 64.0 (45-74) 59.5 (41-74)Gender M/F 5/5 4/6Educational level Lower vocational education Intermediate vocational education Higher vocational education/university

631

352

Employment status Employed Unemployed

37

46

Internet use In daily life For disease purposes

86

97

Location primary tumor Lung Pancreas Jejunum/ileum Unknown

0244

3241

Disease stage (according to UICCb) Stage I/II Stage III Stage IV

019

226

Treatment prior to study participation (%) Somatostatin analogue Curative surgery Palliative surgery Everolimus Radiofrequency ablation Otherc

80%604011

50%412100

Health care use before diagnosis (%) Physical therapist Psychologist

20%11

10%10

a Data are expressed as number unless noted otherwiseb Union for International Cancer Controlc Proton pump inhibitor

3

49


Effect of intervention on outcomesIn the standard care group, one patient was supported by a social worker and another patient by a psychologist during the study period. In the intervention group, two patients received treatment of a physical therapist (n=2) and/or dietician (n=1).Table 2 presents mean scores of the different subscales of EORTC QLQ-INFO25 at baseline and end of study for the two groups and the accompanying effect sizes. Most outcomes on subscales showed negative effect sizes indicating that the intervention group had less improvement in feeling informed. None of the patients wished to receive less information. At baseline, less patients in the intervention group wished to receive more information (50% n the intervention group versus 70% in the standard group).

Baseline and end of study mean scores and accompanying effect sizes of the DT and PL are summarized in Table 3. A medium effect size (ES) was observed for distress and social problems in favor of the intervention group. Trivial or small effect sizes were observed for problems in the other domains. Few patients experienced religious/spiritual problems. Outcomes for QoL measured by EORTC QLQ-C30 and EORTC QLQ-GINET21 questionnaires are depicted in Tables 4A and 4B. The intervention group showed a higher increase in QoL (small ES). A moderate ES was observed for cognitive functioning in favor of the standard care group. In contrast, financial difficulties and pain improved more (moderate effect sizes) in the intervention group. The interven-tion group showed more improvement on disease-specific problems measured by the

Table 2. Perceived information and satisfaction (EORTC QLQ-INFO25)Standard care group (n=10) Intervention group (n=10)

Outcome Pre M (SD) Post M (SD) Pre M (SD) Post M (SD) ES

Information about..

Disease 33.33 (21.52) 42.50 (18.19) 35.83 (21.89) 39.17 (19.66) - 0.30

Medical tests 48.89 (21.72) 56.67 (16.93) 57.78 (15.54) 61.11 (22.38) - 0.22

Treatments 32.78 (13.21) 44.44 (15.27) 29.44 (17.77) 33.33 (16.56) - 0.59

Other services 12.50 (14.83) 30.00 (30.73) 10.83 (15.24) 14.17 (16.22) - 0.74

Different places of care 16.67 (28.32) 23.33 (31.62) 16.67 (28.33) 13.33 (28.11) - 0.17

How to help yourself 13.33 (23.31) 33.33 (31.43) 20.00 (23.31) 20.00 (28.11) - 0.88

Received written information 80.00 (42.16) 80.00 (42.16) 100.00 (0.00) 70.00 (48.30) - 0.52

Received cd/video 0.00 (0.00) 0.00 (0.00) 0.00 (0.00) 0.00 (0.00) 0.00

Satisfaction with information 46.67 (23.31) 50.00 (23.57) 50.00 (23.57) 56.67 (22.50) 0.13

Wanting to receive more info 30.00 (48.30) 60.00 (51.64) 50.00 (52.70) 60.00 (51.64) - 0.28

Wanting to receive less info 100.00 (0.00) 100.00 (0.00) 100.00 (0.00) 100.00 (0.00) 0.00

Helpfullness of information 46.67 (17.21) 56.67 (22.50) 46.67 (17.21) 56.67 (27.44) 0.00

Global Score 38.40 (13.30) 48.08 (12.82) 43.10 (10.51) 43.70 (13.66) - 1.06

Pre M: mean score at baseline, Post M: mean score at 12 weeks, SD: standard deviation, ES: Effect Size.Higher scores means more/better information and satisfaction.

Chapter 3a

50

Table 3. Distress (Distress thermometer and Problem List)Standard care group (n=10) Intervention group (n=10)


Distress level 4.90 (2.96) 4.60 (3.03) 4.10 (3.32) 2.00 (1.49) 0.75

Problems

Practical 5.10 (6.42) 5.60 (6.79) 8.20 (13.30) 7.80 (12.75) 0.16

Social 0.20 (0.63) 0.90 (2.03) 2.60 (5.23) 2.00 (3.27) 0.53

Emotional 17.00 (22.40) 12.50 (14.65) 20.10 (19.20) 12.20 (14.40) 0.20

Spiritual 0.00 (0.00) 0.00 (0.00) 0.50 (1.58) 0.40 (0.97) 0.19

Physical 18.10 (11.83) 11.80 (10.62) 16.90 (11.40) 11.70 (11.41) - 0.10

Global score 40.40 (34.63) 30.80 (27.12) 48.30 (43.33) 34.10 (39.86) 0.18

Pre M: mean score at baseline, Post M: mean score at 12 weeks, SD: standard deviation, ES: effect size.Higher scores mean more distress/problems.

Table 4A. Quality of life (EORTC QLQ-C30)Standard care group (n=10) Intervention group (n=10)


Global quality of life 59.17 (20.95) 58.33 (24.53) 65.83 (26.77) 70.83 (17.67) 0.46

Functioning

Physical 73.33 (23.09) 76.67 (24.19) 79.33 (27.30) 80.67 (20.23) - 0.15

Role 61.67 (28.38) 61.67 (32.44) 66.67 (36.00) 70.00 (21.94) 0.15

Emotional 67.50 (36.74) 73.33 (25.99) 70.00 (32.20) 78.33 (20.11) 0.11

Cognitive 78.33 (27.27) 83.33 (24.85) 78.33 (20.86) 70.00 (23.31) - 0.61

Social 68.33 (14.59) 75.00 (26.35) 73.33 (31.62) 78.33 (20.86) - 0.08

Symptoms

Dyspnoea 23.33 (27.44) 20.00 (23.31) 36.67 (33.15) 30.00 (24.60) 0.15

Insomnia 16.67 (23.57) 20.00 (28.11) 23.33 (27.44) 20.00 (28.11) 0.21

Appetite loss 23.33 (35.31) 10.00 (22.50) 20.00 (35.83) 13.33 (23.31) - 0.30

Constipation 6.67 (14.05) 10.00 (22.50) 16.67 (32.39) 13.33 (23.31) 0.14

Diarrhea 23.33 (22.50) 23.33 (38.65) 16.67 (32.39) 16.67 (32.39) 0.00

Financial difficulties 10.00 (16.10) 23.33 (27.44) 26.67 (43.89) 23.33 (35.31) 0.63

Fatigue 45.56 (31.62) 37.78 (34.03) 33.33 (36.29) 33.33 (28.21) - 0.40

Nausea and vomiting 13.33 (20.49) 1.67 (5.27) 18.33 (33.75) 10.00 (21.08) - 0.12

Pain 26.67 (33.52) 25.00 (26.35) 35.00 (38.05) 21.67 (20.86) 0.52

Pre M: mean score at baseline, Post M: mean score at 12 weeks, SD: standard deviation, ES: effect size.Higher scores for quality of life and functioning represents higher quality of life and level of functioning. Higher scores for symptoms represents a higher level of symptomatology.

3

51


EORTC QLQ-GINET21 questionnaire. For example, large effect sizes were observed for disease related worries. Outcomes of the different domains of empowerment are presented in Table 5. The intervention group felt better informed (moderate ES) and had more confidence in their treatment (small ES). In contrast, the standard care group scored higher for the domain ‘acceptance of illness’ (small ES) and ‘optimism and control over the future’ (moderate ES).

Qualitative and quantitative semi-structured interviewTo evaluate patient’s opinion on the web-based system all patients in the intervention group were interviewed. The number of times the web-based system was visited ranged from 2-5 (n=4 patients), 6-10 (n=2), 11-15 (n=3) to 16-20 (n=1). Frequently visited parts of the web-based sys-tem were ‘diagnostic procedures’ and ‘treatments’. The DT, specific information about

Table 4B. Quality of life (EORTC QLQ-GINET21)Standard care group (n=10) Intervention group (n=10)

Outcome Pre M (SD) N Post M (SD) N Pre M (SD) N Post M (SD) N ES

Endocrine symptoms 27.78 (19.07) 10 22.22 (14.81) 10 25.56 (21.63) 10 13.33 (11.48) 10 0.35

Gastrointestinal symptoms 28.00 (22.40) 10 27.33 (14.55) 10 30.67 (29.18) 10 24.00 (21.36) 10 0.34

Treatment related symptoms 14.44 (18.22) 5 15.54 (18.56) 5 14.44 (18.22) 5 23.33 (16.81) 5 - 0.40

Problems social functioning 28.89 (15.89) 10 35.56 (12.62) 10 42.22 (28.11) 10 38.89 (24.15) 10 0.69

Disease related worries 33.95 (26.42) 9 42.59 (31.18) 9 45.00 (28.50) 10 32.22 (22.03) 10 0.92

Pain muscles/bone 16.67 (23.57) 10 16.67 (23.57) 10 37.50 (45.21) 8 37.50 (33.03) 8 0.00

Problems sexual functioning 9.52 (16.27) 7 14.29 (26.23) 7 38.89 (38.97) 6 33.33 (42.16) 6 0.52

Problems receiving information

13.33 (17.21) 10 20.00 (23.31) 10 23.33 (27.44) 10 10.00 (16.10) 10 1.04

Problems body image 20.00 (17.21) 10 13.33 (17.21) 10 3.33 (10.54) 10 6.67 (14.05) 10 - 0.60

Pre M: mean score at baseline, N: number of patients, Post M: mean score at 12 weeks, SD: standard deviation, ES: effect size.Higher scores means more or worse symptoms/problems.

Table 5. Empowerment (CEO)Standard care group

(n=10)Intervention group

(n=10)

Outcome Mean (SD) Mean (SD) ES

Feeling informed 13.60 (3.86) 15.30 (3.02) 0.51

Confidence in relationship physician 37.06 (6.78) 36.20 (10.17) - 0.10

Confidence in treatment 16.60 (4.35) 18.60 (4.30) 0.46

Acceptance of illness 17.60 (2.12) 16.40 (5.54) - 0.29

Optimism and control over future 25.00 (4.69) 22.50 (3.87) - 0.58

Mean: mean score (at 12 weeks), SD: standard deviation, ES: effect size.Higher scores represent better empowerment for each outcome.

Chapter 3a

52

NET, and the part on ‘diagnostic procedures’ and ‘treatments’ were found most useful. All patients found the web-based system user-friendly. All but one patient found the information presented by the web-based system understandable. Most patients (n=6) found the provided information by the web-based system ‘pretty complete’, ‘complete’ or ‘very complete’. Suggestions for information to add to the web-based system were ‘frequently used medications in NET patients other than tumor treatments (e.g. laxa-tives)’, ‘advice about healthy life style’ and ‘life expectancy’.Seven patients preferred information on their type of tumor only. Two patients were also interested in other types of tumors and enjoyed reading about it. Most patients (n=8) found the web-based system of additional value. All would recommend this web-based system to peers and believed this should be part of the standard care. Also, several patients considered it of interest for people surrounding the NET patient (e.g. family, friends, general practitioners). All wanted to use the web-based system in the future. Patients would like to use it immediately upon diagnosis.

Discussion

This is the first study examining whether use of a web-based information and support system is feasible in patients with NET. The results suggest that use of this web-based system is feasible. The patients with NET appreciated the use of the web-based system and found its use of additional value to standard care. This exploratory study was not powered to detect differences between the two groups. Use of eHealth-tools is increasing and a goal of the European Commission to improve citizens health (20). Given the lack of available eHealth-tools in patients with NET, there is an unmet need for a professionally designed web-based information and support system in these patients. Our results provide information to further develop web-based information systems in rare cancer types like NET. Patients using the web-based information system had less problems with finding ad-equate information and felt better informed as measured by the EORTC QLQ-GINET21 and CEO. Unexpectedly, the results of the EORTC QLQ-INFO25 demonstrated that use of the web-based system resulted in less improvement of feeling informed and satisfaction of received information. The EORTC QLQ-INFO25 is not specifically devel-oped for patients with an NET, which could potentially explain the observed difference between questionnaires. However, it is possible that a web-based system results in feeling worse informed and more depressed. In contrast to face-to-face contact with health care providers, visiting a web-based system does not offer the possibility to discuss feelings of insecurity in person or to ask questions after having read disturbing information about their disease. Also, the intervention group had less acceptance of

3

53


illness and less control and optimism regarding the future at the end of the study period. A randomized controlled study examining different types of internet support groups in breast cancer patients, also reported negative effects on depression and anxiety in the internet support patients who received additional information and support of other breast cancer patients (21). Being provided with more information can possibly be con-fronting for patients with cancer which may result in an increase of negative thoughts and worse future perspectives.This study had a small sample size due to the feasibility design. Younger age and higher educational level are associated with more internet use for health care information (22). Therefore, the intervention group may have had a more critical view on the received in-formation. As a result of these small sample size despite randomization, the intervention group was younger, higher educated and comprised three patients with pulmonary NET. The DT/PL is an element of the web-based system and therefore the intervention group had to fill out the DT/PL online instead of a paper-and-pencil version. Since the content of the online DT/PL is identical to the paper-and-pencil version, we do not expect that this has affected the outcome.The results of the current study have given insight into the existing worries and prob-lems of patients with NET, but also into the needs and expectations of information and support. This has enabled us to develop a sufficiently powered randomized controlled trial investigating an adapted web-based system for patients with both new and a longer known diagnosis of NET (ClinicalTrials.gov Identifier NCT02472678). Mainly based on the results of the semi-structured interview, adaptations have been made to the web-based system, such as adding new information (e.g. information about commonly used medication by patients with NET and average life expectancy), ‘truths and hoaxes’ about NET and experiences of patients with different types of NET with their disease. Hopefully, firm conclusions about effects of this web-based system can be drawn after finishing this study. If the web-based system is proven effective after finishing this study, it will become available for all patients with NET in the Netherlands.

Chapter 3a

54

References 1. Oberg KE. The management of neuroendocrine tumours: current and future medical

therapy options. Clin Oncol (R Coll Radiol) 2012;24:282-93. 2. Ramage JK, Ahmed A, Ardill J, et al. Guidelines for the management of gastroenteropan-

creatic neuroendocrine (including carcinoid) tumours (NETs). Gut 2012;61:6-32. 3. Tuinman MA, Gazendam-Donofrio SM, Hoekstra-Weebers JE. Screening and referral

for psychosocial distress in oncologic practice: use of the Distress Thermometer. Cancer 2008;113:870-8.


5. Frojd C, Larsson G, Lampic C, von Essen L. Health related quality of life and psychosocial function among patients with carcinoid tumours. A longitudinal, prospective, and compara-tive study. Health Qual Life Outcomes 2007;5:18.


7. Davies NJ, Kinman G, Thomas RJ, Bailey T. Information satisfaction in breast and prostate cancer patients: implications for quality of life. Psychooncology 2008;17:1048-52.

8. Cox A, Jenkins V, Catt S, Langridge C, Fallowfield L. Information needs and experiences: an audit of UK cancer patients. Eur J Oncol Nurs 2006;10:263-72.

9. Mallinger JB, Griggs JJ, Shields CG. Patient-centered care and breast cancer survivors‘ satisfaction with information. Patient Educ Couns 2005;57:342-9.

10. Rehnberg G, Absetz P, Aro AR. Women‘s satisfaction with information at breast biopsy in breast cancer screening. Patient Educ Couns 2001;42:1-8.

11. van Uden-Kraan CF, Drossaert CH, Taal E, Shaw BR, Seydel ER, van de Laar MA. Em-powering processes and outcomes of participation in online support groups for patients with breast cancer, arthritis, or fibromyalgia. Qual Health Res 2008;18:405-17.

12. Samoocha D, Bruinvels DJ, Elbers NA, Anema JR, van der Beek AJ. Effectiveness of web-based interventions on patient empowerment: a systematic review and meta-analysis. J Med Internet Res 2010;12:e23.

13. Bouma G, Admiraal JM, de Vries EG, Schroder CP, Walenkamp AM, Reyners AK. Internet-based support programs to alleviate psychosocial and physical symptoms in cancer patients: a literature analysis. Crit Rev Oncol Hematol 2015;95:26-37.

14. Stichting NET-groep, informatie over neuro-endocriene kanker 2017. NET-groep. (2017, at http://www.net-kanker.nl/).

15. Arraras JI, Greimel E, Sezer O, et al. An international validation study of the EORTC QLQ-INFO25 questionnaire: an instrument to assess the information given to cancer patients. Eur J Cancer 2010;46:2726-38.

16. Aaronson NK, Ahmedzai S, Bergman B, et al. The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 1993;85:365-76.


3

55


18. van Uden-Kraan CF, Drossaert CH, Taal E, Seydel ER, van de Laar MA. Participation in online patient support groups endorses patients’ empowerment. Patient Educ Couns 2009;74:61-9.

19. Manuals 2018. EORTC. (at http://groups.eortc.be/qol/manuals). 20. 2018. European commission. (at https://ec.europa.eu/health/ehealth/overview_en). 21. Lepore SJ, Buzaglo JS, Lieberman MA, Golant M, Greener JR, Davey A. Comparing stan-

dard versus prosocial internet support groups for patients with breast cancer: a randomized controlled trial of the helper therapy principle. J Clin Oncol 2014;32:4081-6.

22. van de Poll-Franse LV, van Eenbergen MC. Internet use by cancer survivors: current use and future wishes. Support Care Cancer 2008;16:1189-95.

Chapter 3a

56

Supplementary data.

Extensive description study proceduresEligible patients were informed by their treating medical oncologist about the study dur-ing a follow-up visit or by phone if the next follow-up visit was planned beyond 2 months. Patients who were interested in study participation were subsequently approached by one of the investigators and received study information. Following informed consent, patients were randomized in a 1:1 ratio to standard care group or the intervention group that received standard care with additional access to the web-based system. Random-ization was performed by e-mail from the investigator to the central data manager of the Department of Medical Oncology (UMCG) who is not involved in the recruitment or treatment of the patients. For allocation to the standard care or intervention group a computer-generated randomization list was used. The allocation sequence was con-cealed from the investigator. The investigator was informed by e-mail to which group the patient was assigned. The investigator informed the patient by phone about the randomization outcome and further process. Also, the treating medical oncologist was notified about the randomization outcome. After randomization, all patients received questionnaires by mail about socio-demographic characteristics, health care/internet use, perception and satisfaction of received information (EORTC QLQ-INFO25) and quality of life (EORTC QLQ-C30, EORTC QLQ-GINET21) as baseline measurement. The standard care group also received the paper-and-pencil version of the Distress Thermometer (DT) and Problem List (PL). After the investigator had collected these completed questionnaires, the intervention group received via their preferred route an e-mail or letter with a short introduction of the web-based system together with log-in information including a personal username and password. These patients were asked to complete the online DT and PL within one week after having received the log-in infor-mation (baseline measure DT and PL). At week 12, all patients received questionnaires by mail about health care use, perception and satisfaction of received information, qual-ity of life and empowerment (Constructs Empowering Outcomes-questionnaire) as end of study measurement. The standard care group also received the paper-and-pencil version of the DT and PL. The intervention group were reminded by e-mail/letter to complete the online version of the DT and PL after 12 weeks. Patients assigned to the intervention group were also requested to complete a semi-structured interview regarding their opinion on the web-based system by phone.

3b

Web-based personalized information and support for patients with a

neuroendocrine tumor: randomized controlled trial

L.D. de Hosson‡1, G. Bouma‡1, J. Stelwagen1, H. van Essen1, G.H. de Bock2, D.J.A. de Groot1, E.G.E. de Vries1, A.M.E. Walenkamp1

‡ The first two authors contributed equally to the manuscript

1Department of Medical Oncology, University of Groningen, University Medical Center Groningen, Groningen, the Netherlands.

2Department of Epidemiology, University of Groningen, University Medical Center Groningen, Groningen, the Netherlands.

Chapter 3b

60

Abstract

Background Patients with a neuroendocrine tumor (NET) frequently have physical and psychoso-cial complaints. Aim of this study is to determine whether a web-based, personalized information and support system (WINS) reduces distress and/or improves patients’ perception of and satisfaction with information received and quality of life (QoL).

Methods Patients with NET, stratified for those newly diagnosed (<6 months, n=28) and with a longer history of disease (n=74), were randomized between standard care (n=49) and intervention, consisting of access to WINS (n=53). Primary outcome was distress and satisfaction with perceived information measured with the distress thermometer and problem list and the QOL questionnaire (QLQ)-INFO25. Secondary outcomes were quality of life (QoL) and empowerment, measured with QLQ-C30 and QLQ-GINET21, and the construct of empowerment (CEO) questionnaire. The intervention group also completed a questionnaire based on the technical acceptance model.

Results We observed no difference in distress slope and slope of median global score on per-ceived information and satisfaction between the intervention and control group. Also, QoL did not improve. Empowerment was better in the control group for 4 of 5 constructs. Interestingly, 55% of patients wished to receive more information at baseline.

Conclusions In a population of patients with NET, access to WINS did not improve indicators for distress, perception of information and satisfaction with information received, more than standard care only. Despite the need for more information, the WINS does not have added value to the information and care provided by health care professionals. Clinical Trial Registration ClinicalTrials.gov (NCT02472678)

Keywords neuroendocrine tumor, information, quality of life, internet, web-based system

3

61

Web-based personalized information and support for patients with a neuroendocrine tumor

Background and aims

Neuroendocrine tumors (NETs) are rare tumors with an incidence of 3.5/100,000 per year in the last decade (1,2). Patients with NET may experience various symptoms from the tumor mass, the output of hormones secreted by the tumor, treatment and accompanying side effects (3). Patients are frequently metastasized at the time of diagnosis. Patients with metastasized NETs have a relatively long median survival: 100 months for NETs of the small intestine and 60 months for pancreatic NETs (4). Patients with NET have a lower quality of life (QoL) compared with the general population (5). In cancer survivors, receiving adequate information is associated with health-related QoL and is seen as an essential aspect of supportive care(6). For patients with NET it is difficult to find meaningful and understandable information about their diagnosis. In an observational study of NET patients, using qualitative interviews, 7 out of 18 patients found the internet to be a useful source of general information (7). We previously devel-oped a web-based personalized information and support system (WINS), for patients with NET with the aim of reducing distress and/or improving patients’ perception of and satisfaction with information received. A pilot study demonstrated the feasibility of using WINS in patients with NET (8). Based on these results and on patients’ recommenda-tions, we developed the current version of WINS, which was used in the present study. The aim of this randomized trial is to determine whether WINS reduces distress and/or improves NETs patients’ perception of and satisfaction with information received, QoL and empowerment.

Methods

ParticipantsEligible participants were adult patients treated at the University Medical Centre Gron-ingen (UMCG) Department of Medical Oncology for NET grade 1 or 2 (World Health Organization 2010 classification) with the primary tumor at any site of origin, and who were proficient in Dutch (both reading and writing). Patients with a life expectancy of less than 3 months as evaluated by their doctor were excluded. The study was ap-proved by the medical ethical committee of the UMCG and registered in ClinicalTrials.gov (NCT02472678). All patients gave written informed consent.

Randomization process and study procedureFor a detailed description of study procedures, see Supplementary Information and Figure 1. The included patients were stratified randomized for those diagnosed within 6 months and those with disease duration ≥6 months. Patients were randomized 1:1

Chapter 3b

62

to the control group, receiving standard care, or the intervention group, which received standard care with additional access to WINS. At baseline patients’ socio-demographic and disease characteristics, internet use and health care use were collected. Patients in both groups received a questionnaire about their perception of and satisfaction with the received information and their QoL. The control group was also given a question-naire about distress and problems. After returning the questionnaires, patients in the intervention group received log-in information for website access. At the first website visit, the patients were asked to complete a questionnaire about distress and problems at baseline, before they were given access to the other items on the website. At 12 weeks follow-up, all patients were asked to complete the questionnaires again, with an additional questionnaire to asses empowerment. Patients in the intervention group were asked to complete the questionnaire about distress and problems at the WINS instead of completing it with pen and paper. Furthermore, they were requested to complete an additional questionnaire about their use of and opinion about WINS.

Standard care and study interventionAll patients received standard care. At their first visit at the Department of Medical Oncology patients are informed verbally about the disease and they meet the oncology nurse to become acquainted. During follow-up visits, the medical oncologist evaluates the general well-being, discusses possible treatment options and treatment side-effects (if applicable) and answers questions of the patient. Between follow-up visits all patients could consult an oncology nurse 24 hours a day, 7 days a week by telephone. Further-more, patients were able to get information from the NET patient association (9). If physical and/or psychosocial problems require more in-depth discussion, investigation or treatment, patients can receive a consultation with the oncology nurse, oncologist or other health care professionals.In the intervention group, the questionnaire about distress and problems served as a self-screening tool for physical and psychosocial problems. By using WINS, patients could obtain personalized information about reported problems and, if necessary, request professional care. Patients could send an e-mail with a question or request a telephone consultation with the researcher (a physician experienced in treating NET patients) in case of questions, problems or a request for referral. On the WINS site, pa-tients could also find general information about the disease, read about the experiences of other NET patients and find links to other websites about QoL issues not specifically related to the disease.

Outcome measurementsIllness-related patient characteristics were extracted from the medical records at base-line. All other measurements were performed by self-report questionnaires.

3

63


Distress was measured using the validated Dutch Distress Thermometer (DT) and Problem list (PL) (10). This questionnaire consisted of one single item that asks patients to indicate the amount of overall distress experienced during the past week and a PL with several items divided into five domains. Higher scores indicate more distress or more problems. The EORTC QLQ-INFO25 in Dutch was used to evaluate patients’ perception of and satisfaction with information received (11). Higher scores indicate better perceived information provision. QoL was measured by the cancer-specific Dutch EORTC QLQ-C30 (version 3.0) and the NET-specific EORTC QLQ-GINET21 (12,13). Higher scores indicate higher QoL at the functional and global health/QoL scale and higher symptom burden for the symptom scales of the QLQs (12). Empowerment was measured with specific domains of the ‘Constructs Empowering Outcomes’ (CEO) ques-tionnaire (14). Higher scores indicate greater empowerment (14). This questionnaire has a retrospective nature. Furthermore, this questionnaire has not been extensively validated. Consequently, our findings should be interpreted with caution (14-16). An additional questionnaire, only given to the intervention group at end of study was based on the constructs of ‘perceived usefulness’ and ‘attitude and usage’ from the revised Technology Acceptance Model and on a self-constructed question on a 5-point likert scale; if patients recommend WINS to other NET patients (17).

Sample size calculationSample size calculation was based on the results of the pilot study, in which only newly diagnosed patients were included. To detect a significant difference in the change of the distress thermometer between the control and intervention group, using an independent t-test with an effect size of 0.6, we calculated that 90 patients had to be included (Sup-portive Information). We used a per-protocol analysis.

Statistical analysis For descriptive statistics, mean and standard deviation (sd) for normal distribution and median and interquartile range for other distributions and frequencies were calculated for all measures. The scores of the EORTC questionnaires were calculated according to the EORTC guidelines (11-13). An independent t-test or Mann Whitney U test was per-formed, depending on the kind of distribution, to detect differences between the effect of standard care complemented with the WINS versus standard care alone. Subgroup analysis was performed in newly diagnosed patients. Differences were considered sig-nificant at p <0.05. Analyses were performed with the software package SPSS, version 23 for Windows (SPSS, Inc, Chicago, IL, USA).

Chapter 3b

64

Results

PatientsBetween May 2015 and October 2016, we included 105 patients in the study (Figure 1). The trial was ended when 91 patients completed the study. Of the 91 patients who com-pleted the study, 46 (12 newly diagnosed) patients were randomized to the intervention group. Baseline characteristics of included patients are shown in Table 1. At baseline, 78 patients had previously visited the internet for information on NET before random-ization. At baseline, 50 patients stated that they wanted to receive more information compared to 31 at end of study. At baseline, 4 patients answered that they would prefer to receive less information. At the end of study no patients gave this answer.

Primary outcomesThe median distress level in both groups was 3 before and after study (Table 2). A significant difference in distress was found for only the domain ‘social problems’; at end of study 12 patients in the control group of 45, reported social problems compared to 7 patients at baseline. Four patients at end of study in the intervention group of 46 patients, compared to 10 patients at baseline were found to have social problems (p<0.01). The median global score for patients’ perception of and satisfaction with information received, did not improve in the intervention group relative to the control group (Table 3). Interestingly, 53 and 57% of patients in the control and intervention group wish to receive more information, respectively. After the intervention less patients

82

Assessed for eligibility (n=156)

Allocated to control group (n=51)

Randomized (n=105)

Does not meet +inclusion criteria (n= 1) Declined to participate (n= 50) Not accustomed to the internet or IT (n=27) Did not want to be reminded of their disease (n=3) Too stressful to participate (n=5) No reason (n=15)

Lost to follow-up (n=2) Discontinued intervention because patient did not wanted to be reminded of their disease (n=2) Died (n=1) Screening failure (n=1)

Allocated to intervention group (n= 54)

Analyzed (n=45)

Lost to follow-up (n=4) No longer able to complete questionnaire due to disease progression (n=1) Died (n=2) Screening failure (n=1)

Analyzed (n=46)

Figure 1. Consort diagram

Figure 1. Consort diagram

3

65


in the control group (31%) wished to receive more information versus the patients in the intervention group (38%).

Table 1. Baseline characteristics

All patientsn=91

Control groupn=45

Intervention groupn=46

Newly diagnosed patients n (%) 25 (27) 13 (29) 12 (26)

Sex (female) n (%) 42 (46) 23 (51) 19 (41)

Age mean in years (sd) 62 ( 8) 63 (7) 62 (10)

Location of primary tumor: n (%)

Pancreas 21 (23) 9 (20) 12 (26)

Intestine 53 (58) 26 (58) 27 (59)

Stomach/duodenum 3 (3) 2 (4) 1 (2)

Colorectal 2 (3) 2 (4) 1 (2)

Bronchopulmonal 1 (1) 1 (2) 0 (0)

Appendix 1 (1) 1 (2) 0 (0)

Unknown/other 9 (10) 4 (9) 5 (11)

Duration of disease in months: mean (SD) 41 (56) 33 (38) 50 (66)

Disease grade: n (%)

1 57 (63) 27 (60) 30 (65)

2 21 (23) 12 (27) 9 (20)

Unknown 13 (14) 6 (13) 7 (15)

Marital state (Married) n (%) 75 (82) 34 (76) 41 (89)

Education (Polytechnic or university) n (%) 33 (37) 16 (36) 17 (38)

Internet use n (%)

Daily 86 (96) 40 (91) 46 (100)

For information about the disease 78 (87) 37 (82) 41 (89)

Treatment during study: n (%)

Surgery 7 (8) 2 (4) 5 (11)

PRRT 2 (2) 1 (2) 1 (2)

SSA 64 (70) 32 (71) 32 (70)

Systemic treatment other than SSA 17 (19) 9 (20) 8 (17)

Treatment before study: n (%)

Surgery 52 (57) 26 (58) 26 (57)

PRRT 2 (2) 0 (0) 2 (4)

Radiotherapy 3 (3) 0 (0) 3 (6)

SSA 65 (71) 32 (71) 33 (72)

Systemic treatment other than SSA 23 (25) 12 (26) 11 (24)

Other (RFA) 1 (1) 0 (0) 1 (2)

n= number, PRRT= peptide receptor radionucleide therapy, RFA= radiofrequency ablation, SSA= somatostatin analogue

Chapter 3b

66

Secondary outcomesNo significant differences were found between the control and intervention groups for each symptom, problem, or level of functioning regarding QoL (Table 4).

Table 2. Distress (Distress thermometer and Problem List)Control group (n=45) Intervention group (n=46)

OutcomePreMedian (range)

PostMedian (range)

PreMedian (range)

PostMedian (range) p

Distress level (0-10) 3 (1-5) 3 (1-5) 3 (2-5) 3 (1-5) NS

Practical problems 0 (0-4) 0 (0-4) 0 (0-4) 0 (0-5) NS

Social problems 0 (0-0) 0 (0-4) 0 (0-0) 0 (0-0) 0.002

Emotional problems 3.5 (0-13) 6 (0-12) 7 (1-16) 7 (1-16) NS

Spiritual problems 0 (0-0) 0 (0-0) 0 (0-0) 0 (0-0) NS

Physical problems 8 (4-18) 10 (6-19) 13 (7-21) 12 (5-20) NS

Global score 14 (7-32) 17 (6-44) 21 (12-41) 21 (7-44) NS

Higher scores mean more distress (from problems).NS= no significant difference, Pre Median: median score at baseline, Post median; median score a 12 weeks, range; interquartile range.

Table 3. Perceived information and satisfaction (EORTC QLQ-INFO25)Control group (n=45) Intervention group (n=46)


PostMedian (range)

PreMedian (range)

PostMedian (range)

p

Information about

Disease 50 (33-58) 50 (42-65) 50 (33-60) 50 (33-67) NS

Medical tests 67 (44-67) 67 (44-67) 67 (56-67) 67 (28-56) NS

Treatments 44 (25-56) 39 (33-56) 44 (28-56) 8 (0-27) NS

Other services 17 (0-25) 17 (8-33) 17 (6-23) 8 (0-27) NS

Different location of care facilities

0 (0-33) 33 (0-33) 0 (0-33) 0 (0-33) NS

How to help yourself 33 (0-33) 33 (0-33) 17 (0-33) 33 (0-33) NS

Satisfaction with information 67 (33-67) 67 (33-67) 67 (33-67) 67 (33-67) NS

Helpfulness of information 67 (50-67) 67 (67-67) 67 (67-67) 67 (33-67) NS

Percentage of patients

Received written information 91 91 78 80

Received cd/video 2 2 2 0

Wish to receive more info 53 31 57 38

Wish to receive less info 2 0 2 0

Outcome Pre M (SD) Post M (SD) Pre M (SD) Post M (SD) p

Global Score 49 (37-55) 51 (42-59) 45 (33-56) 38 (47-57) NS

Higher scores indicate more/better information and satisfaction.cd= compact disk, NS= no significant difference, Pre Median= median score at baseline, Post median= median score a 12 weeks, range= interquartile range.

3

67


Table 4A. Quality of life (EORTC QLQ-C30)Control group (n=45) Intervention group (n=46)


PostMedian (range)

PreMedian (range)


Global quality of life 75 (58-83) 75 (62-83) 75 (67-83) 75 (65-83) NS

Physical 87 (67-100) 87 (67-97) 87 (73-100) 87 (72-100) NS

Role 83 (50-100) 83 (67-100) 83 (63-100) 83 (67-100) NS

Emotional 92 (75-100) 83 (67-96) 83 (67-96) 83 (67-94) NS

Cognitive 100 (67-100) 83 (67-100) 100 (83-100) 100 (67-100) NS

Social 100 (67-100) 100 (83-100) 100 (67-100) 83 (67-100) NS

Dyspnea 0 (0-0) 0 (0-33) 0 (0-0) 0 (0-0) NS

Insomnia 0 (0-33) 33 (0-33) 33 (0-33) 33 (0-33) NS

Appetite loss 0 (0-33) 0 (0-0) 0 (0-0) 0 (0-0) NS

Constipation 0 (0-0) 0 (0-0) 0 (0-0) 0 (0-0) NS

Diarrhea 0 (0-33) 0 (0-50) 0 (0-33) 0 (0-0) NS

Financial difficulties 0 (0-0) 0 (0-0) 0 (0-0) 0 (0-0) NS

Fatigue 33 (17-56) 22 (11-44) 22 (11-39) 28 (11-44) NS

Nausea and vomiting 0 (0-17) 0 (0-8) 0 (0-0) 0 (0-0) NS

Pain 0 (0-33) 17 (0-33) 17 (0-33) 17 (0-33) NS

Higher scores for quality of life and functioning indicate higher quality of life and level of functioning. Higher scores for symptoms indicate greater severity. NS= no significant difference, Pre Median= median score at baseline, Post median=median score a 12 weeks, range= interquartile range.

Table 4B. Quality of life (EORTC QLQ-GINET21)Control group (n=45) Intervention group (n=46)


PostMedian (range)

PreMedian (range)


Endocrine symptoms 11 (0-33) 11 (0-33) 0 (0-14) 11 (0-33) NS

Gastrointestinal symptoms 20 (7-27) 20 (12-33) 20 (13-28) 20 (7-33) NS

Treatment related symptoms 11 (0-22) 17 (0-28) 11 (0-22) 0 (0-22) NS

Problems social functioning 33 (17-44) 22 (11-33) 33 (22-56) 22 (11-44) NS

Disease related worries 33 (19-53) 33 (17-50) 44 (22-67) 33 (19-44) NS

Pain muscles/bone 33 (0-33) 33 (0-33) 33 (0-67) 33 (0-33) NS

Problems sexual functioning 0 (0-50) 0 (0-33) 17 (0-33) 0 (0-67) NS

Problems receiving information 0 (0-0) 0 (0-0) 0 (0-33) 0 (0-0) NS

Problems body image 0 (0-33) 0 (0-33) 0 (0-33) 0 (0-0) NS

Higher scores indicate more or worse symptoms/problems.NS= no significant difference, Pre Median: median score at baseline, Post median; median score a 12 weeks, range; interquartile range.

Chapter 3b

68

Empowerment was better in the control group, for all constructs, except for ‘optimism and control over future’ for which no difference was found (Table 5). Most patients agreed with the statements mentioned in the additional questionnaire (Table 6). During the study, the median number of visits to the website was 3 (range 2-4), and only 3 patients used the opportunity to ask questions and consult with the researcher for clinical purposes. Other questions were about technical or logistical aspects.

Newly diagnosed patientsThe planned subgroup analysis for newly diagnosed patients did not detect any differ-ence in this subgroup regarding the global score for distress and the problem domains (Tables S2 and S3).In this subgroup, perceived information and satisfaction did not differ between the con-trol group and intervention group. We found a difference for only one item, ‘information about the disease in the QLQ-INFO25’; after 12 weeks this score decreased in the in-tervention group and increased in the control group (p=0.046). Furthermore, regarding the secondary endpoints, QoL and empowerment, most items did not show significant

Table 5. Empowerment (CEO)Control group (n=45) Intervention group (n=46)

Outcome Median (range) Median (range) p

Feeling informed 16 (14-16) 13 (12-14) 0.019

Confidence in relationship physician 37 (34-41) 32 (26-37) <0.001

Confidence in treatment 19 (16-20) 15 (12-18) <0.001

Acceptance of illness 18 (15-20) 15 (11-17) 0.001

Optimism and control over future 24 (22-26) 23 (16-25) NS

Higher scores represent better empowerment for each outcome.NS= no significant difference, Median: median score (at 12 weeks), range; interquartile range.

Table 6. Questionnaire based on patients’ opinion and use of the website (based on con-structs of the Technology Acceptance Model)


Outcome Median (range)

The website is useful to me 4 (4-4)

The information at the website is interesting to me 4 (4-4)

I find this a site that adds value 4 (3-5)

I have a positive attitude towards the website 4 (4-5)

I would recommend the site to peers 4 (3-5)

How often do you visit the website 3 (2-4)

Higher scores indicate more agreement with the statement (except for number of visits) Median: median score (at 12 weeks), range; interquartile range.

3

69


differences were found between the control and intervention groups. Most patients agreed with the statements on the self-constructed questionnaire (Tables S4-S6).

Discussion

In this randomized trial we found that WINS did not reduce distress nor improved NETs patients’ perception of and satisfaction with information received, or QoL. Empower-ment was better in the control group. We found the same results in the pre-planned subgroup analyses with newly diagnosed patients. Developing a WBS providing patient detailed information corresponding to their individ-ual needs and wishes is difficult (18). The contact and communication with health care professionals remains a crucial source of information and support for patients with NET. Other web-based studies have shown that web-based technology does not replace patient-provider communication (19). In one such study, in which 103 cancer patients received questionnaires about the WBS, the highest rated component was receiving an answer from a nurse. In another trial in 766 patients receiving chemotherapy for solid cancers, patients were randomized to either standard care or to standard care plus a web-based self-reporting system. Physicians received symptom printouts at visits and nurses received e-mail alerts when participants of the intervention group reported severe or worsening symptoms. In this trial, QoL improved (34% vs. 18%; p<0.001) and OS was higher (31 vs. 26 months; p=0.03) in the intervention group compared to the group receiving standard care only (20). In a randomized controlled study of a website providing additional information and symptom self-management support in 325 breast and prostate cancer patients, distress was used as the primary endpoint (21). Compared to the control group, which was given URLs of publically available cancer information websites, an improvement for patients in the intervention group was found only on the subscale ‘global distress index’. A potential explanation for the unexpected negative results of our study could be that NET patients do not need additional information. However, at baseline half of the pa-tients in both groups indicated that they wanted more information about the disease. During study period, the need for more information decreased in both groups. A decline in the need for information during the course of the disease was observed in other studies analyzing patients with other types of cancer (22,23). In our trial all patients had easy access to specialists and nurses during standard care, which might have satisfied their need for information. Furthermore, patients were able to get information from the NET patient association (9). Patients only visited the WBS sporadically; this sporadic behavior was also reported in other trials using a WBS (21,24). In a qualitative interview study, one of the three main reasons reported for not using a WBS was that patients

Chapter 3b

70

had sufficient access to information elsewhere (25). In two other studies, in patients with cancer and brain tumors, patients reported that they avoided using the WBS because they did not want to be reminded of their suffering or, because they had a preference for other kinds of communication (25,26). In our study, empowerment was better in the control group. In other trials in cancer patients no statistically differences were seen in empowerment between the control group and the intervention group with access to the web-based intervention (27,28) Due to the indolent natural course of NET, and because NET patients may not have symptoms at study inclusion, it is challenging to demonstrate an improvement in QoL in this population (4). In the RADIANT-4, a large RCT on the use of everolimus in advanced non-functional, gastrointestinal or lung NET patients, few patients had symp-toms at study inclusion. During the study period, global QoL remained stable in both groups, as also seen in our study (29). Despite its well-powered prospective randomized controlled design, our study also has some limitations, in particular the use of the CEO questionnaire. Due to the ret-rospective nature of the CEO questionnaire, no baseline scores could be determined. Furthermore, this questionnaire has not been extensively validated. Consequently, our finding that empowerment at the end of the study was better in the control group should be interpreted with caution (14-16). Another limitation is that we used a per-protocol analysis. Because the primary endpoint was measured with questionnaires, patients who withdrew early did not complete the end-of-study questionnaires. Therefore an intention-to-treat analysis was not possible in this trial. Probably, even, with an intention-to-treat analysis, no significant differences could be shown.

ConclusionIn NET patients, WINS did not improve distress scores and patients’ perception of or satisfaction with information received, compared to patients receiving standard care only. Despite the need for more information, the WINS does not have added value to the information and care provided by health care professionals.

3

71


References 1. Ramage JK, Ahmed A, Ardill J, et al. Guidelines for the management of gastroenteropan-

creatic neuroendocrine (including carcinoid) tumours (NETs). Gut 2012;61:6-32. 2. Lawrence B, Gustafsson BI, Chan A, Svejda B, Kidd M, Modlin IM. The epidemiology of gas-

troenteropancreatic neuroendocrine tumors. Endocrinol Metab Clin North Am 2011;40:1,18, vii.

3. Oberg KE. The management of neuroendocrine tumours: current and future medical therapy options. Clin Oncol (R Coll Radiol) 2012;24:282-93.

4. Dasari A, Shen C, Halperin D, et al. Trends in the Incidence, Prevalence, and Survival Outcomes in Patients With Neuroendocrine Tumors in the United States. JAMA Oncol 2017;3:1335-42.




8. Bouma G, de Hosson LD, van Woerkom CE, et al. Web-based information and support for patients with a newly diagnosed neuroendocrine tumor: a feasibility study. Support Care Cancer 2017;25:2075-83.

9. Stichting NET-groep, informatie over neuro-endocriene kanker 2017. NET-groep. (2017, at http://www.net-kanker.nl/).

10. Tuinman MA, Gazendam-Donofrio SM, Hoekstra-Weebers JE. Screening and referral for psychosocial distress in oncologic practice: use of the Distress Thermometer. Cancer 2008;113:870-8.

11. Arraras JI, Greimel E, Sezer O, et al. An international validation study of the EORTC QLQ-INFO25 questionnaire: an instrument to assess the information given to cancer patients. Eur J Cancer 2010;46:2726-38.

12. Aaronson NK, Ahmedzai S, Bergman B, et al. The European Organization for Research and Treatment of Cancer QLQ-C30: a quality-of-life instrument for use in international clinical trials in oncology. J Natl Cancer Inst 1993;85:365-76.


14. van Uden-Kraan CF, Drossaert CH, Taal E, Shaw BR, Seydel ER, van de Laar MA. Em-powering processes and outcomes of participation in online support groups for patients with breast cancer, arthritis, or fibromyalgia. Qual Health Res 2008;18:405-17.

15. Admiraal JM, van der Velden AWG, Geerling JI, et al. Web-based tailored psycho-education for breast cancer patients at the onset of the survivorship phase: a multicenter randomized controlled trial. J Pain Symptom Manage 2017;54:466-475.

16. van Uden-Kraan CF, Drossaert CH, Taal E, Seydel ER, van de Laar MA. Participation in online patient support groups endorses patients’ empowerment. Patient Educ Couns 2009;74:61-9.

Chapter 3b

72

17. Heijden van der. Factors influencing the usage of wesites: the case of a generic portal in the Netherlands. Information & Management 2003;40:541.

18. Kuenzel U, Monga Sindeu T, Schroth S, Huebner J, Herth N. Evaluation of the quality of on-line information for patients with rare cancers: Thyroid cancer. J Cancer Educ 2017;:Epub ahead of print.

19. Ruland CM, Maffei RM, Borosund E, Krahn A, Andersen T, Grimsbo GH. Evaluation of different features of an eHealth application for personalized illness management support: cancer patients’ use and appraisal of usefulness. Int J Med Inform 2013;82:593-603.

20. Basch E, Deal AM, Dueck AC, et al. Overall survival results of a trial assessing patient-reported outcomes for symptom monitoring during routine cancer treatment. JAMA 2017;318:197-8.

21. Ruland CM, Andersen T, Jeneson A, et al. Effects of an internet support system to assist cancer patients in reducing symptom distress: a randomized controlled trial. Cancer Nurs 2013;36:6-17.

22. S. Mattson, E.M.G. Olsson, B. Johansson, M. Carlsson,. Health-related internet use in people with cancer: results from a cross-sectional study in two outpatient clinics in Sweden. 2017;:e163.

23. Matsuyama RK, Kuhn LA, Molisani A, Wilson-Genderson MC. Cancer patients’ information needs the first nine months after diagnosis. Patient Educ Couns 2013;90:96-102.

24. Vogel RI, Petzel SV, Cragg J, et al. Development and pilot of an advance care planning website for women with ovarian cancer: a randomized controlled trial. Gynecol Oncol 2013;131:430-6.

25. Varsi C, Gammon D, Wibe T, Ruland CM. Patients’ reported reasons for non-use of an internet-based patient-provider communication service: qualitative interview study. J Med Internet Res 2013;15:e246.

26. Piil K, Jakobsen J, Juhler M, Jarden M. The feasibility of a brain tumour website. Eur J Oncol Nurs 2015;19:686-93.

27. Admiraal JM, van der Velden AWG, Geerling JI, et al. Web-Based tailored psychoeducation for breast cancer patients at the onset of the survivorship phase: A multicenter randomized controlled trial. J Pain Symptom Manage 2017;54:466-75.

28. van den Berg SW, Gielissen MF, Custers JA, van der Graaf WT, Ottevanger PB, Prins JB. BREATH: Web-based self-management for psychological adjustment after primary breast cancer--results of a multicenter randomized controlled trial. J Clin Oncol 2015;33:2763-71.

29. Pavel ME, Singh S, Strosberg JR, et al. Health-related quality of life for everolimus versus placebo in patients with advanced, non-functional, well-differentiated gastrointestinal or lung neuroendocrine tumours (RADIANT-4): a multicentre, randomised, double-blind, placebo-controlled, phase 3 trial. Lancet Oncol 2017;18:1411-22.

3

73


Supplementary Table S1. Questionnaire based on patients’ opinion and use of the website (based on constructs of the Technology Acceptance Model)









Higher scores indicate more agreement with the statement (except for number of visits) Median: median score (at 12 weeks), range; interquartile range.

Supplementary Table S2. Distress in newly diagnosed patients (Distress thermometer and Problem List)

Control group (n=13) Intervention group (n=12)


PostMedian (range)

PreMedian (range)


Distress level (0-10) 4 (2-6) 5 (2-7) 3 (0-3) 3 (0-3) NS

Practical problems 0 (0-3) 0 (0-5) 0 (0-0) 0 (0-2) NS

Social problems 0 (0-0) 0 (0-10) 0 (0-0) 0 (0-0) NS

Emotional problems 0 (4-12) 0 (0-16) 3 (0-15) 3 (0-8) NS

Spiritual problems 0 (0-2) 0 (0-0) 0 (0-0) 0 (0-0) NS

Physical problems 12 (7-18) 14 (5-24) 8 (4-14) 11 (2-15) NS

Global score 15 (13-33) 35 (9-57) 16 (7-40) 14 (3-27) NS

Higher scores indicate more distress (from problems).NS= no significant difference, Pre Median: median score at baseline, Post median: median score at 12 weeks, range: interquartile range.

Chapter 3b

74

Supplementary Table S3. Perceived information and satisfaction with information in newly diagnosed patients (EORTC QLQ-INFO25)

Control group (n=13) Intervention group (n=12)


PostMedian (range)

PreMedian (range)


Information about

Disease 42 (38-63) 58 (46-67) 50 (27-73) 38 (25-67) 0.046

Medical tests 67 (50-67) 67 (56-67) 56 (33-67) 67 (44-67) NS

Treatments 33 (22-53) 39 (31-61) 42 (24-54) 40 (22-61) NS

Other services 4 (17-29) 25 (4-50) 13 (0-25) 8 (0-37) NS

Different location of care facilities

0 (0-33) 25 (4-50) 0 (0-33) 0 (0-58) NS

How to help yourself 33 (0-33) 33 (0-50) 33 (0-33) 33 (8-58) NS

Satisfaction with information 67 (33-67) 67 (50-67) 67 (33-67) 67 (33-100) NS

Helpfulness of information 67 (50-67) 67 (67-67) 67 (67-67) 67 (33-100) NS

Percentage of patients

Received written information 100 100 100 82

Received cd/video 0 8 0 0

Wish to receive more info 54 15 58 64

Wish to receive less info 0 0 0 0

Outcome Pre M (SD) Post M (SD) Pre M (SD) Post M (SD) p

Global Score 52 (41-57) 55 (52-62) 48 (39-56) 42 (31-63) NS

Higher score indicates more/better information and satisfaction.cd: compact disk, NS: no significant difference, Pre Median: median score at baseline, Post median” median score a 12 weeks, range: interquartile range.

3

75


Supplementary Table S4A. Quality of life in newly diagnosed patients (EORTC QLQ-C30)Control group (n=13) Intervention group (n=12)


PostMedian (range)

PreMedian (range)


Global quality of life 75 (50-83) 75 (54-83) 75 (69-83) 75 (67-83) NS

Physical 80 (53-100) 80 (53-93) 93 (87-100) 87 (82-98) NS

Role 67 (42092) 67 (67-100) 83 (67-100) 83 (53-100) NS

Emotional 83 (75-100) 83 (63-96) 79 (67-92) 79 (67-100) NS

Cognitive 67 (67-100) 67 (67-92) 100 (83-100) 75 (67-100) NS

Social 83 (50-100) 83 (67-100) 100 (83-100) 83 (67-100) NS

Dyspnea 0 (0-33) 0 (0-50) 0 (0-0) 0 (0-0) NS

Insomnia 0 (0-33) 33 (17-33) 33 (0-33) 33 (0-33) NS

Appetite loss 0 (0-33) 0 (0-0) 0 (0-0) 0 (0-0) NS

Constipation 0 (0-0) 0 (0-0) 0 (0-0) 0 (0-0) NS

Diarrhea 0 (0-33) 33 (0-67) 0 (0-33) 33 (0-58) NS

Financial difficulties 0 (0-0) 0 (0-0) 0 (0-0) 0 (0-0) NS

Fatigue 44 (6-56) 33 (22-44) 17 (3-31) 22 (11-61) NS

Nausea and vomiting 0 (0-8) 0 (0-17) 0 (0-0) 0 (0-0) NS

Pain 33 (0-75) 33 (8-42) 8 (0-29) 0 (0-33) NS

Higher scores for quality of life and functioning indicate higher quality of life and level of functioning. Higher scores for symptoms indicate greater severity.NS: no significant difference, Pre Median: median score at baseline, Post median: median score a 12 weeks, range: interquartile range.

Supplementary Table S4B. Quality of life in newly diagnosed patients (EORTC QLQ-GINET21)Control group (n=13) Intervention group (n=12)


PostMedian (range)

PreMedian (range)


Endocrine symptoms 11 (0-33) 0 (0-19) 0 (0-19) 0 (0-8) NS

Gastrointestinal symptoms 20 (3-40) 27 (9 -37) 17 (13-20) 17 (13-32) NS

Treatment related symptoms 16 (6-28) 22 (6-33) 14 (0-22) 0 (0-22) NS

Problems social functioning 33 (28-67) 33 (17-61) 33 (25-53) 33 (22-53) NS

Disease-related worries 33 (19-72) 33 (25-72) 53 (36-67) 33 (17-44) 0.006

Pain muscles/bone 33 (0-50) 33 (8-58) 0 (0-25) 0 (0-33) NS

Problems sexual functioning 0 (0-58) 17 (0-75) 0 (0-33) 0 (0-42) NS

Problems receiving information 0 (0-33) 0 (0-0) 0 (0-33) 0 (0-0) NS

Problems body image 0 (0-33) 0 (0-33) 0 (0-0) 0 (0-0) NS

Higher scores indicate more or worse symptoms/problems.NS: no significant difference, Pre Median: median score at baseline, Post median: median score at 12 weeks, range: interquartile range.

Chapter 3b

76

Supplementary Table S5. Empowerment in newly diagnosed patients (CEO)Control group (n=13) Intervention group (n=12)

Outcome Median (range) Median (range) p

Feeling informed 16 (14-18) 14 (13-16) NS

Confidence in relationship physician 34 (33-42) 31 (25-39) 0.04

Confidence in treatment 19 (16-20) 16 (10-19) NS

Acceptance of illness 17 (15-20) 13 (9-20) NS

Optimism and control over future 24 (23-28) 16 (14-26) NS

Higher scores indicate better empowerment for each outcome.NS: no significant difference, Median: median score (at 12 weeks), range: interquartile range.

Supplementary Table S6. Questionnaire on patients’ opinions and use of the website in newly diagnosed patients (based on constructs of the Technology Acceptance Model)









Higher scores indicate more agreement with the statement except for number of visits Median: median score (at 12 weeks), range: interquartile range.

3

77


Supplementary Information

Randomization process and study procedureThe included patients were stratified randomized for those diagnosed within 6 months and those with disease duration ≥6 months. We expected that newly diagnosed patients would have other information needs which could lead to bias if not stratified. Patients were randomized 1:1 to the control group, receiving standard care, or the intervention group, which received standard care with additional access to WINS. Randomization was performed by the central data manager of the Department of Medical Oncology (UMCG) who was not involved in other aspects of the study. A computer-generated randomization list was used for allocation to the control or intervention arm. The al-location sequence was concealed from the investigator. The investigator was informed by e-mail to which arm the patient was assigned. The investigator informed the patient by phone about the randomization outcome and further process. At baseline patients’ socio-demographic and disease characteristics, internet use and health care use were collected. Patients in both groups received a questionnaire about their perception of and satisfaction with the received information and their QoL. The control group was also given a questionnaire about distress and problems. After returning the questionnaires, patients in the intervention group received log-in information for website access. At the first website visit, the patients were asked to complete a questionnaire (within 1 week) about distress and problems at baseline, before they were given access to the other items on the website. At 12 weeks follow-up, all patients were asked to complete the questionnaires again, with an additional questionnaire to asses empowerment. Patients in the intervention group were asked to complete the questionnaire about distress and problems at the WINS instead of completing it with pen and paper. Furthermore they were requested to complete an additional questionnaire about their use of and opinion about WINS.

Sample size calculationSample size calculation was based on the results of the pilot study, in which only newly diagnosed patients were included. The distress score had a Cohen’s d effect size of 0.75 in favor of the intervention group and an effect size of 1.0 was found in favor of the control group for the EORTC QLQ-INFO25. Since we expected using a WINS for newly diagnosed patients and for patients diagnosed more than 6 months before inclusion, might have a smaller effect, than for only newly diagnosed patients, we adapted the effect size to 0.6 for the distress thermometer and to 0.8 for the EORTC QLQ-INFO 25. To detect a significant difference in the change of the distress thermometer between the control and intervention group, using an independent t-test with an effect size of 0.6, we calculated that 90 patients had to be included. We used a per-protocol analysis.

4

Towards optimal personalized diet and vitamin supplementation in NET

patients

L.D. de Hosson1, J. Stelwagen1, G. Bouma1, B. Sijtema2, S, Huitema2, H.J.R. van Faassen3, G.H. de Bock4, D.J.A. de Groot1, M.J.E. Campmans-Kuijpers5, I.P. Kema3, E.G.E. de Vries1, A.M.E. Walenkamp1

1Department of Medical Oncology, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands.

2Dieticians, Department of Medical Oncology, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands

3Department of Laboratory Medicine, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands.

4Department of Epidemiology, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands

5Department of Gastroenterology, University Medical Center Groningen, University of Groningen, Groningen, The Netherlands

Chapter 4

80

Keywords nutrition, neuroendocrine tumor, somatostatin analogue, vitamin deficiency, diet

4

81

Towards optimal personalized diet and vitamin supplementation in NET patients

Letter

Patients with a neuroendocrine tumor (NET) often have gastro-intestinal complaints, leading to impaired absorption of critical food components such as vitamins. The aim of this feasibility study is to explore if ‘personalized dietary advice and supplementa-tion of deficient vitamins’ (DIVIT) is feasible in NET-patients using a SSA to perform a well-powered clinical trial. In this feasibility study, fifteen NET patients using a SSA for over 6 months got the intervention DIVIT. Feasibility was assessed by calculation of participation/dropout rate and number of intervention-related adverse events. Results showed a participation rate of 75% and dropout rate of 17%. No intervention-related adverse events occurred. We conclude that DIVIT is feasible in NET patients using a SSA. Results from this study will support the development of a well-powered clinical trial.Patients with neuroendocrine tumors often have gastro-intestinal complaints (1). Previ-ously, fat-soluble vitamin deficiencies during long-term somatostatin analogue (SSA) treatment were demonstrated (2). Deficient vitamins can lead to complaints. An inverse correlation between vitamin D levels and overall survival (OS) has been observed in 138 gastro-enteropancreatic NET patients (3). NET-patients experience a lower quality of life (QoL) as compared with the general population, amongst others, due to increased bowel movements (1). Strikingly little is known about how NET patients using a SSA should be best supported with dietary advice or supplementation of vitamins. The aim of this feasibility study is to explore if DIVIT is feasible in NET-patients using a SSA. Results from this study will support the development of a well-powered clinical trial. Eligible participants were adult patients with NET grade 1 or 2, treated at the University Medical Center Groningen (UMCG), receiving a SSA for more than 6 months, and were proficient in Dutch. Patients with serotonin-producing tumors and with non-serotonin producing tumors were included. Patients with a life expectancy of less than 3 months, a history of another malignancy or planned for abdominal surgery were excluded. The study was approved by the medical ethical committee of the UMCG and registered in ClinicalTrials.gov (NCT02481804). All patients gave written informed consent. Measurements were performed at baseline (T0), week 4 (T4) and end of study (T18). During the first 4 weeks, standard care was continued. Thereafter patients started with DIVIT. Dietician consultations were conducted at T4 by 1 out-patient visit, and after 5, 10 and 15 weeks by telephone. Vitamin supplementation started at T4 or at week 8. The intervention, DIVIT, was based on the patients’ individual situation and on Dutch Guidelines and provided patients insight in how they could adjust their diet to obtain a healthy diet leading to less gastro-intestinal symptoms (4). For a vitamin plasma vitamin K1 < 0.8 nmol/L, oral phytomenadione once daily 2 mg during 2 months was

Chapter 4

82

supplied (5). For a plasma vitamin A < 0.8 μmol/L, retinol 25,000 IU once daily during 2 weeks, followed by 5000 IU once daily was supplied, until sufficient values were reached. Patients with a plasma 25-OHvit D < 50 nmol/L were treated with 800 IU mg cholecalciferol once daily, chronically. Furthermore, if patients had plasma vitamin E < 19.2 μmol/L, alpha-tocopherol acetate 100 mg once daily was supplied, until sufficient values. In case patients had tryptophan in platelet rich plasma <40 μmol/L, nicotinamide once daily 250 mg was supplied. Vitamin B12 was supplied if plasma levels were < 145 pmol/L with a start dose of hydroxocobalamin 1000 mcg 2 times a week intramuscular during 1 month, followed by 1000 mcg once in 2 months (6).The primary outcome; feasibility, was assessed by calculation of participation/dropout rate and number of intervention-related adverse events. At baseline patient- and illness-related characteristics were collected by examination of the medical records. QoL was measured with the European Organization for Research and Treatment (EORTC) questionnaire for gastro-intestinal NET (QLQ-GINET21) and for cancer pa-tients the QLQ-C30. General distress and distress from problems were measured with the distress thermometer and problem list (PL). Empowerment was measured with four domains of the Dutch version of the Construct Empowering Outcomes (CEO) question-naire. The Patient-Generated Subjective Global Assessment (PG-SGA), was used to measure patients’ nutritional state.Vitamin A, D, E, K1, B12 in plasma were analyzed. Tryptophan levels in patients and serotonin levels for routine analysis were assessed in samples of platelet rich plasma (7). 24-hour Urine was used for analysis of 5-hydroxyindolacetic acid and N1-methylnic-otinamide (N1-MN), the main metabolite of vitamin B3, (8). Mean and standard deviation for normal distribution and median and interquartile range for other distributions were calculated. The scores of questionnaires were calculated according to corresponding guidelines. Mean scores were used for each domains of the PL and CEO questionnaire. Effect sizes (ES) for difference in scores on each endpoint of the EORTC questionnaires, PL and PG-SGA, proportion of patients with normal vitamin values, and ES of the outcomes of CEO were determined with Cohen’s d. An ES of 0.20-0.50 is defined as small, between 0.50-0.80 as medium and ≥ 0.80 as large. Analyses were performed with SPSS, version 23 for windows (SPSS, Inc, Chicago, IL, USA).Twenty-four patients with NET were invited to participate between May 2015 and Octo-ber 2015. Eighteen patients were included and fifteen completed the full study, giving a participation rate of 75% and a dropout rate of 17%. Baseline characteristics of patients that completed the study are shown in Table 1. Hospital admission occurred in three patients for non-study related adverse events.

4

83


Table 1. Baseline characteristics n = 15Characteristic

Age – median (range) 61 (56-69)

Sex n (%)

Male 7 (47)

Female 8 (53)

Location primary tumor n (%)

Lung 2 (13)

Stomach 1 (7)

Jejunum/ileum 9 (60)

Unknown 3 (20)

Serotonin producing1 n (%)

Yes 13 (85)

Tumor grade n (%)

Grade 1 10 (67)

Grade 2 3 (20)

Unknown 2 (13)

Prior antitumor treatment n (%)

Somatostatin analogue 15 (100)

Palliative surgery 10 (67)

Abdominal surgery2 9 (60)

Interferon 2 (13)

Everolimus 1 (7)

Radiotherapy 1 (7)

Prior use of diet n (%)

Without consultation of a dietician 2 (13)

With consultation of a dietician; 3 (20)

Energy enriched diet 2 (13)

Diet to without large vessels to prevent intestinal obstruction 1 (7)

Prior vitamin use n (%)

Total patients with vitamin use 6 (40)

Vitamin D 4 (27)

Vitamin K 3 (20)

Vitamin B12 2 (13)

Nicotinamide 3 (20)

M: mean score, n: number of patients, SD: standard deviation.1. The NET was defined as serotonin-producing if platelet rich plasma serotonin level > 5.4 nmol/ 109 platelets or if 5-HIAA in 24/hour urine > 3.8 mmol/mmol creatinine at baseline measurement.2. Two patients with abdominal surgery had a hemicolectomy, one patient had a hemicolectomy and sigmoid resection, one patient had a hemicolectomy and a partial small bowel resection and 4 patients had a small bowel resection.

Chapter 4

84

All patients got one or more personalized dietary advices. Nine patients used all di-etary advices. One patient said he did not need additional advices. Eleven patients got supplementation of vitamins during the intervention period.No or only a small effect was found on most items of QoL as measured by the EORTC QLQ-GINET21 and EORTC QLQ-C30. Only the item ‘body image’ of the QLQ-GINET21 worsened after intervention (ES=0.71). ‘Pain’, improved after intervention with a me-dium effect size and emotional functioning was negatively affected by DIVIT (ES=0.59). The distress thermometer and PL showed an increase in practical problems (ES=1.17) and a decrease in physical problems (ES=0.79) during intervention. Furthermore the empowerment decreased on the domain ‘relation with physician’ (ES=0.58 ), ‘feel better informed’ (ES=1.04) and ‘confidence in treatment‘ (ES=0.98). A negligible effect (ES=0.15) was seen on nutrition state measured with the PG-SGA. At start of interven-tion, nine patients had vitamin deficiencies. Despite DIVIT, which improved vitamin levels in five patients, seven patients had deficient vitamin levels after the intervention (ES=0.13) (Figure 1). This is the first prospective feasibility study examining whether a personalized dietary intervention with supplementation of deficient vitamins is feasible in SSA-using patients with NET. Several vitamin deficiencies were present at baseline, indicating that standard care might be insufficient. Despite the addition of DIVIT for 14 weeks, which improved vitamin levels in some patients, almost half of patients had one or more deficient vitamin levels after the intervention. This can be due to malabsorption, or continuous need instead of temporary use of vitamins. Vitamin supplementation was studied before, like in cystic fibrosis (CF), where patients are at risk for fat-soluble vitamin deficiencies. The restoration of serum vitamin to the normal range after supplementation with vitamins was reported (5). The doses of vitamin A, E, K and B12 supplementation used in our study were similar with the doses used in other trials; only vitamin K was supplemented during 8 weeks in our study instead of life long. The vitamin D dose used in our study was based on the required dose for adults without adequate sun exposure and higher than the daily recommended dose by the Institute of Medicine. However, patients with vitamin D deficiency due to malabsorption probably need higher doses (9). Based on our results in combination with the rapid metabolic turnover of vitamin K, we concluded that probably a prolonged duration of vitamin K supplementation and a higher dose of vitamin D supplementation might lead to restoration of vitamin levels in more patients.In NET an association between vitamin D supplementation and OS (p < 0.002) was found (3). Furthermore our intervention showed minimal effect on QoL, as is seen more often in trials measuring the effect of vitamin supplementation in patients with other diseases (5). With the results of the current study it became possible to design a well-powered clinical trial investigating DIVIT for SSA-using patients with NET (NCT03143946). Primary end-

4

85


Figure 1. Vitamin values over timeEach dot represents the vitamin value of a patient at the specific time point. The line shows the lower reference value for the specific vitamin.

106

Vitamin A

Time (weeks)

Vita

min

A (u

mol

/L)

0 4 180

2

4

6

Vitamin D

Time (weeks)

Vita

min

D (n

mol

/L)

0 4 180

50

100

150

107

Vitamin E

Time (weeks)

Vita

min

E (u

mol

/L)

0 4 180

20

40

60

80

Vitamin K

Time (weeks)

Vita

min

K (n

mol

/L)

0 4 180

5

10

15

108

VItamin B12

Time (weeks)

Vita

min

B 1

2 (p

mol

/L)

0 4 180

500

1000

1500

2000

Niacin

Time (weeks)

Nia

cin

(um

ol/L

)

0 4 180

100

200

300

400

Figure 1. Vitamin values over time

Each dot represents the vitamin value of a patient at the specific time point. The line shows the lower reference value for the specific vitamin.

108

VItamin B12

Time (weeks)

Vita

min

B 1

2 (p

mol

/L)

0 4 180

500

1000

1500

2000

Niacin

Time (weeks)

Nia

cin

(um

ol/L

)

0 4 180

100

200

300

400

Figure 1. Vitamin values over time

Each dot represents the vitamin value of a patient at the specific time point. The line shows the lower reference value for the specific vitamin.

Chapter 4

86

point of this single arm intervention trial is the proportion of patients with normal vitamin values, measured with quantitative analysis of blood and urine after the intervention In conclusion, DIVIT is feasible in SSA-using patients with NET. SSA-using patients with NET are at risk for vitamin deficiencies, which are not easily corrected by dietary advices or standard vitamin supplementation. With the results of the current study it became possible to design a well-powered clinical trial to determine if an intervention with DIVIT leads to a change in proportion of patients with normal vitamin values.

4

87


References 1. Pearman TP, Beaumont JL, Cella D, Neary MP, Yao J. Health-related quality of life in pa-

tients with neuroendocrine tumors: an investigation of treatment type, disease status, and symptom burden. Support Care Cancer 2016.

2. Fiebrich HB, Van Den Berg G, Kema IP, et al. Deficiencies in fat-soluble vitamins in long-term users of somatostatin analogue. Aliment Pharmacol Ther 2010;32:1398-404.

3. Massironi S, Zilli A, Bernasconi S, et al. Impact of vitamin D on the clinical outcome of gastro-entero-pancreatic neuroendocrine neoplasms: Report on a series from a single institute. Neuroendocrinology 2017;105:403-11.

4. Gezondheidsraad. Richtlijnen goede voeding. 2006. 5. Jagannath VA, Fedorowicz Z, Thaker V, Chang AB. Vitamin K supplementation for cystic

fibrosis. Cochrane Database Syst Rev 2015;1:CD008482. 6. Aanbevelingen voor vitamines, mineralen, en spoorelementen. Factsheet 2014. Brink ea.

(at https://issuu.com/voedingscentrum/docs/factsheet_aanbevelingen_voor_vitami/2?ff=true&e=1222161/30846004).

7. Kema IP, Meijer WG, Meiborg G, Ooms B, Willemse PH, de Vries EG. Profiling of tryptophan-related plasma indoles in patients with carcinoid tumors by automated, on-line, solid-phase extraction and HPLC with fluorescence detection. Clin Chem 2001;47:1811-20.

8. Bouma G, van Faassen M, Kats-Ugurlu G, de Vries EG, Kema IP, Walenkamp AM. Niacin (Vitamin B3) supplementation in patients with serotonin-producing neuroendocrine tumor. Neuroendocrinology 2016;103:489-94.

9. Lo CW, Paris PW, Clemens TL, Nolan J, Holick MF. Vitamin D absorption in healthy subjects and in patients with intestinal malabsorption syndromes. Am J Clin Nutr 1985;42:644-9.

5

Interlesional heterogeneity of metastatic neuroendocrine tumors

based on [18F]FDOPA PET/CT

Lotte D. de Hosson1, Aline M. van der Loo- van der Schaaf2, Ronald Boellaard3, Johannes H. van Snick3, Elisabeth G.E. de Vries1, Adrienne H. Brouwers3, Annemiek M.E. Walenkamp1

1Department of Medical Oncology, University of Groningen, University Medical Center Groningen, Groningen, The Netherlands

2Department of Radiology, University Medical Center Groningen, University of Groningen, Groningen. Current address, Medical Center Zuiderzee, Lelystad, The Netherlands

3Department of Nuclear Medicine and Molecular Imaging, University of Groningen, University Medical Center Groningen, Groningen, The Netherlands

Chapter 5

90

Abstract

PurposeNeuroendocrine tumors (NETs) can produce neuroendocrine amines resulting in symptoms. Selecting the most active amine-producing tumor lesions for local treat-ment might be beneficial for patients with metastatic small intestinal (SI-)NET. Tumor burden, correlates with catecholamine pathway activity. We analyzed interlesional heterogeneity with 6-[18F]fluoro-L-3,4-dihydroxyphenylalanine ([18F]FDOPA) Positron Emission Tomography (PET) scans in patients with SI-NET and investigated if lesions with substantially higher [18F]FDOPA uptake could be identified.

ProceduresFor all lesions on [18F]FDOPA PET scans the [18F]FDOPA uptake was calculated by dividing standard uptake value (SUV) peak of the lesion by the SUV mean of the background organ. The magnitude of heterogeneity between lesions within a patient was calculated by dividing the lesion with highest by the one with lowest [18F]FDOPA uptake. Lesions with a higher [18F]FDOPA uptake than the upper inner or outer fence (>1.5 or 3 times the interquartile range above the third quartile) were defined as lesions with mild or extreme high [18F]FDOPA uptake, respectively and presence of these were determined in patients with ≥10 lesions.

ResultsWith [18F]FDOPA, over 680 lesions in 38 patients, of which 35 serotonin producing, were detected. [18F]FDOPA uptake varied with a median of 8-fold up to 44-fold between lesions within a patient. In 12 of 20 evaluable patients lesions with mild high [18F]FDOPA uptake were found and in five, lesions with extreme high [18F]FDOPA uptake.

Conclusions[18F]FDOPA PET showed considerable heterogeneity in [18F]FDOPA uptake between tu-mor lesions and identified lesions within patients with mild or extreme high [18F]FDOPA uptake.

Key Words: Neuroendocrine tumor, [18F]FDOPA PET scan, heterogeneity, tracer uptake

5

91

Interlesional heterogeneity of metastatic neuroendocrine tumors based on [18F]FDOPA PET/CT

Introduction

Neuroendocrine tumors of small intestine (SI-NET) are rare tumors which in general are characterized by slow tumor growth. Over 50% of patients with NET present with metastatic disease. Symptoms are caused by tumor mass and/or (over-) production of neuroendocrine amines such as serotonin (1). Heterogeneity of tumor characteristics is increasingly appreciated and common in NETs and occurs at different genetic and epigenetic levels and in the expression of protein biomarkers (2). In several pathology studies heterogeneity between tumor lesions is shown. Immunohistochemical analysis of surgical specimens of liver metastases of 29 patients with pancreatic NET demon-strated heterogeneity within and between synchronous and metachronous metastases for Mindbomb E3 ubiquitin protein ligase 1, CD34, and somatostatin receptor 2 (SS-TRA2) (3). In another study, in 14 out of 27 patients with SI-NET a discordant tumor grade between liver lesions was demonstrated (4). This was also shown in pancreatic NET where 10 out of 16 patients had a discordant tumor grade between nodal, liver or other metastases (5). Furthermore eight out of 26 patients had tumor lesions with no to weak SSTRA2 expression and lesions with moderate to strong SSTRA2 expres-sion (6). For patients with SI-NET a number of systemic and local treatment options are available. Selection of the most active tumor lesion for local treatment might be beneficial for patients with SI-NET to reduce symptoms caused by overproduction of neuroendocrine amines. Local treatment options consists of radiofrequency ablation (RFA), transarterial chemo-embolization (TACE), transarterial embolization (TAE), resection, or the more recently developed, selective internal radiation therapy (SIRT) (7,8). In patients with colorectal liver metastases SIRT added to standard chemotherapy had a better median PFS in the liver than chemotherapy (9). Currently, no studies ana-lyzing heterogeneity between tumor lesions based on molecular imaging are available. In patients with SI-NET the tumor can be visualized with several molecular imaging techniques. [18F]FDOPA visualizes the metabolism of catecholamine pathway in the tumor cell. Total tumor burden measured with [18F]FDOPA PET scans in patients with SI-NET correlates with tumor markers of the serotonin and catecholamine pathway (10). We analyzed interlesional tumor [18F]FDOPA uptake heterogeneity with correction for background activity with PET scans and investigated if tumor lesions with substantially higher [18F]FDOPA uptake could be identified in patients with SI-NET.

Chapter 5

92

Materials and methods

Patients Patient records of all patients with NET undergoing an [18F]FDOPA PET scan at the University Medical Center of Groningen (UMCG), The Netherlands, between February 2014 until April 2015, were screened. All adult patients with a metastasized SI-NET with visualization of more than one lesion on the [18F]FDOPA PET scan and who had under-gone a diagnostic computed tomographic (CT) scan within 6 months of the PET scan were selected for further analysis (Fig. S1). The following baseline characteristics were retrieved from patients medical record; duration of disease, previous treatments, tumor grade (1 or 2) according to the 2010 WHO classification, 5-HIAA level in 24-h urine and serotonin level in platelet rich plasma. A tumor was considered serotonin-producing if either a 5-HIAA level in 24-h urine >3.8 mmol/mol creatinine or a serotonin level >5.4 nmol/109 thrombocytes in platelet rich plasma was measured (11). [18F]FDOPA PET scans were conducted during standard care. Because of the retrospective nature of this analysis, according to the Dutch regulations and the ethical committee of our institution, no approval by this committee was needed and no additional informed consent was required. The patient whose scan was used in Figure 2, gave informed consent for publication, according to author guidelines.

[18F]FDOPA PET and CT Scan[18F]FDOPA was produced in the radiochemical laboratory of the UMCG as described previously (12). Patients fasted for 6 hours before the tracer injection and were allowed to continue all medication. For the reduction of tracer decarboxylation and subsequent renal clearance, all patients received 2 mg/kg carbidopa (maximum 150 mg) orally as pre-treatment, 1 hour before the [18F]FDOPA injection to increase [18F]FDOPA uptake in tumor cells (13). PET images were acquired 60 min after intravenous administration of [18F]FDOPA (170-215 MBq). Scanning was performed from the upper legs to head with a PET-CT camera (Siemens Biograph mCT 40 or 64 slices, 4 detector rings) with zoom factor 1. Scanning time per bed position was 1 to 3 min depending on body weight, and a low dose CT was used for attenuation and scatter correction. PET data were reconstructed with Siemens Ultra HD (trueX and time of flight), and ordered subsets expectation maximization reconstruction (OSEM), using 3 iterations and 21 subsets and a matrix of 400 with a full width at half maximum (FWHM) of 5 mm Gaussian (isotropic) filter (10,14,15). The PET scan was combined, at the same time or within 6 months before or after date of PET scanning, with a contrast enhanced diagnostic CT scan. A CT scan of the chest (n=32) and abdomen (n=38) was obtained, with iodine-containing intravenous, Iomeron 350 mg/mL, and oral contrast agents. CT scans were

5

93


performed on the Biograph mCT PET-CT, 40-slices, or 64-slices, Somatom Sensation 64, or Somatom Definition Dual (all Siemens) with a maximal slice thickness of 2 mm.

Imaging Data Analysis PET-positive tumor lesions were defined as lesions with an unequivocal visibility on [18F]FDOPA PET scan greater than normal activity in that body region (16). Around all PET-positive tumor lesions a volume of interest (VOI) was manually drawn on the PET scan with use of in-house developed Accurate Software. In each ‘background organ’, a spherical VOI was drawn in an area of homogenous activity in the organ, to quantify the physiological [18F]FDOPA uptake. The VOI diameter used to measure SUV mean in background organs was 1 cm in blood pool, bone marrow at L4 or L5, cortical bone of the humerus, cortex of kidney and pancreas; the VOI diameter was 2 cm in muscle, myocardium and spleen; and the VOI diameter was 3 cm in brain, liver and lung. SUV max, SUV peak and SUV mean values were calculated according to the body weight of the patient.To detect and quantify all tumor lesions, including PET-negative tumor lesions, [18F]FDOPA PET scans were compared with CT scans that were reviewed by a radiologist with knowledge of the PET data. The radiologist counted in each organ all abnormalities and defined these as definitely benign, definitely malignant, or as ‘inconclusive lesions’. Afterwards, all discrepancies between malignant lesions or inconclusive lesions on CT scan versus tumor lesions on [18F]FDOPA PET scans were reviewed by one of the investigators. If a PET-negative, definitely malignant or ‘inconclusive lesion’ was found, a VOI was drawn surrounding the lesion on CT scan to calculate SUV values of the same region on the [18F]FDOPA PET scans. For small PET-negative lesions <0.15 mL, like lung nodules, a spherical VOI with a diameter of 1 cm was drawn. In case of more than 20 liver lesions on [18F]FDOPA PET or CT scan, a cut-off of 20 lesions for both modalities was used. If > 20 liver lesions were present, single lesions could not be identified separately and therefore could not be counted exactly. PET-negative, CT defi-nitely malignant lesions were included for further analysis of [18F]FDOPA heterogeneity. Lesions that were ‘inconclusive’ on CT and PET-negative were not used for further analysis of [18F]FDOPA heterogeneity. Patients were evaluable for detection of lesions with mild or extreme high [18F]FDOPA uptake if > 10 tumor lesions were present, see for calculation below. [18F]FDOPA uptake in a tumor lesion was calculated by dividing the SUV peak value of the tumor lesion by the SUV mean value measured in the VOI drawn in the background organ. To determine the ‘tumor burden’ for each lesion or VOI the uptake has to be multiplied by lesion or VOI volume (10,17). This method was analyzed in a study using 18fluorodeoxyglucose-PET scans. SUV peak corrected for local SUV mean background multiplied by volume had the most optimal feasibility and repeatability to measure metabolic active tumor

Chapter 5

94

volume (i.e. tumor burden) (17). To explore the partial volume effect of the results, SUVs of tumor lesions were correlated with tumor VOI volume and the correlation was compared with the relation between SUV and volume seen in the NEMA NU 2-2007 phantom model (18). To explore effect of tissue density on SUV, the SUV of background organs was correlated with tissue density. Tissue density was based on the average Hounsfield Units per VOI.

Statistical AnalysisPatients’ demographic and clinical variables were summarized using medians with ranges for continuous variables and frequencies with percentages for categorical variables. The magnitude of heterogeneity between lesions within a patient was calcu-lated by dividing the lesion with highest [18F]FDOPA uptake with the lesion with lowest [18F]FDOPA uptake within a patient. Differences between a median [18F]FDOPA uptake or SUV mean per organ was analyzed with the Kruskal Wallis test. To determine lesions with mild or extreme higher [18F]FDOPA uptake than other lesions, the lower (Q1) and upper quartiles (Q3) were determined first. The interquartile range (IQR), or Q3–Q1, was then computed. Lastly, the upper fences were computed as follows: upper inner fence Q3 +1.5* (IQR) and upper outer fence Q3 +3 * (IQR). Any data points outside of the fences, are considered as lesions with mild or extreme high [18F]FDOPA uptake, compared to the other lesions in that patient. Correlations of parametric values were calculated using Pearson’s r, nonparametric values were calculated using Spearman’s rho. A p-value < 0.05 was considered significant. Statistical analyses were performed using SPSS version 23 (SPSS, Inc., Chicago, IL).

Results

PatientsIn the selected period 182 [18F]FDOPA PET scans of 160 patients with SI-NET were performed. Eighty three patients were excluded because their scan showed no or only one tumor lesion, 29 patients had a primary NET of pancreatic or lung origin and in 10 patients the time between [18F]FDOPA PET scan and CT scan was more than 6 months (Fig. S1). In total 38 patients were eligible for analysis. Data of 5-HIAA urinalysis and serotonin in platelet rich plasma were available for 34 and 38 patients, respectively. According to these data, 35 patients had a serotonin producing NET. Baseline charac-teristics of included patients are shown in Table 1. Median age of the patients was 65 years (interquartile range 55-70), and 55% of patients were male.

5

95


Analysis of tumor lesions measured with [18F]FDOPA PET Over 680 tumor lesions were visualized on the [18F]FDOPA PET scans (mean > 20 lesions per patient). The median SUV peak in tumor lesions was 5.5 (interquartile range 3.1-11.2) and the median [18F]FDOPA uptake was 4.1 (interquartile range 2.2-7.7). In total 581 [18F]FDOPA PET lesions could be correlated to a morphologic lesion on CT scan (Fig. S2). Thirty eight [18F]FDOPA PET-negative lesions were detected by CT scan. In one patient the CT scan revealed two liver metastases that were not detected on [18F]FDOPA PET scan and these were included for analysis of [18F]FDOPA hetero-geneity. Of 36 other [18F]FDOPA PET-negative lesions detected by CT scan, mostly localized in lung and lymph nodes, no differentiation could be made on CT whether the lesions were benign or malignant and these were not analyzed. In some patients one PET lesion referred to two or more CT lesions, because the lesions were merged with each other on [18F]FDOPA PET scan. Seven lesions were seen on the PET-scan but outside the field of view on CT-scan.Examination of the lesions per organ showed 91 bone lesions visible on [18F]FDOPA PET scan, inside the view of field of the CT scan, but not visible on CT scan. When we evaluated other organs than the skeleton, a small number of lesions were revealed only on [18F]FDOPA PET scan (liver, lymph node/mesenterium, lung) or only on CT scan (lung) (Fig. S2).

Interlesional heterogeneity and physiological [18F]FDOPA uptakeThe [18F]FDOPA uptake ranged from 1- to 44-fold between individual lesions within the same patient (median 8-fold) (Table 2). Twelve of the 20 patients with over 10 tumor lesions had lesions with mild high [18F]FDOPA uptake and five patients had lesions with extreme high [18F]FDOPA uptake (Fig. 1,2). Liver, lymph nodes combined with

Table 1. Baseline characteristics of small intestinal NET-patiens (n=38)Characteristic n of patients (%)

Sex male, 21 (55)

Tumor grade:12Unknown

27 (71)3 (8)8 (21)

Treatment: SSA use Surgery Everolimus Interferon PRRT Any treatment

23 (61)22 (58)1 (3)3 (8)1 (3)26 (68)

Patients with serotonin producing-tumor 35 (92)

n= number, NET=neuroendocrine tumors, PRRT=peptide radionuclide receptor therapy, SD=standard deviation, SSA=somatostatin analogue.

Chapter 5

96

Table 2. Tracer uptake, magnitude of heterogeneity and number of lesions with mild or ex-treme high tracer uptake.

Characteristics of metastases Tracer uptakePt Lesion (n) Metastasized organs Mean Max vs

min MH (n)*

EH (n) *

1 >65 Lung, Liver, Pancreas, LN, Bone, Other 6.5 43.6 3 1

2 17 Liver 2.2 3.7 0 0

3 3 Liver, LN 1.5 1.2 - -

4 38 Lung, Liver, LN 6.0 3.6 0 0

5 6 Bowel, LN 3.0 3.0 - -

6 5 Lung, Liver 3.4 1.8 - -

7 >32 Liver, LN 6.6 9.5 0 0

8 9 Liver, Pancreas, Bowel, LN, Other 3.3 5.3 - -

9 14 Liver 3.9 8.5 1 0

10 13 Lung, Liver, LN 7.0 22.1 0 0

11 17 Liver, Bowel, LN 2.7 5.6 2 1

12 5 LN 8.1 15.0 - -

13 7 Heart, Pancreas, Bowel, LN, Other 5.7 6.2 - -

14 9 Liver, LN 2.5 2.2 - -

15 38 Lung, Liver, LN, Bone, Other 12.4 42.3 2 0

16 12 Lung, Pancreas, LN, Bone 6.1 3.8 0 0

17 2 Bowel, LN 9.5 12.5 - -

18 >40 Liver, Bowel, LN, Bone, Other 7.5 30.7 3 1

19 9 Liver, Bowel, LN 10.6 28.7 - -

20 6 LN, Bone 13.7 28.0 - -

21 5 Liver, LN 3.5 8.4 - -

22 4 Pancreas, LN 10.5 8.3 - -

23 46 Liver, Pancreas, LN, Bone 7.8 25.0 1 0

24 17 Liver, LN 3.8 8.6 1 0

25 13 Liver, Pancreas, Other 3.5 5.0 0 0

26 20 Liver, Pancreas, LN, Bone 5.3 10.9 0 0

27 >30 Heart, Liver, LN, Bone, Other 7.7 13.9 4 0

28 >23 Liver, LN 5.8 11.2 0 0

29 4 Bowel, LN 5.2 2.9 - -

30 7 Liver 1.6 1.4 - -

31 3 LN 2.8 2.1 - -

32 5 LN 4.0 2.1 - -

33 >54 Lung, Liver, LN, Bone, Other 10.3 171 3 0

34 5 Heart, Bowel, LN 5.5 6.2 - -

35 15 Liver, LN 3.1 6.5 1 0

36 8 Liver, Bowel, LN 4.0 9.1 - -

37 15 Liver, Bone 6.2 17.2 1 1

38 >48 Liver, LN, Bone, Other 2.8 8.4 9 7

LN = lymph node, Max vs min = lesion with highest tracer uptake divided by lesion with lowest tracer uptake within a patient, MH= lesion with mild higher [18F]FDOPA uptake than other lesions, EH= lesion with extreme higher [18F]FDOPA uptake than other lesions n= number.

5

97


mesenterium, bone and lungs were most frequently involved. Tumor burden per lesion is depicted in the Fig. S3. The median [18F]FDOPA uptake of all analyzed lesions dif-fered per organ (p<0.05) (Fig. 3). Lesions in lung had the highest [18F]FDOPA uptake (median 8.4) and lesions of heart the lowest (median 1.7). The median of SUV mean values measured in the background organs of all patients was 1.2 (interquartile range of 0.9-1.5) and differed per organ (p<0.05 ) (Fig. 4 and Table 3).

Effect of partial volume effect and tissue density on physiological [18F]FDOPA uptakeSUV peak and [18F]FDOPA uptake of all tumor lesions were correlated with their vol-umes, r=0.20 (p= 7.33*10-8) and r=0.15 (p=6.7*10-5), respectively (Fig 5). SUV mean of healthy lung was linearly correlated with tissue density r=0.83 (p=1.28*10-10) (Fig 6, Table 3).

Figure 1. [18F]FDOPA uptake per tumor lesion per patient. Tumor lesions above the inner and outer upper fences show mild or extreme high [18F]FDOPA uptake in patients with ≥ 10 tumor lesions. Each symbol represents a single lesion. Color and kind of symbol are equal per lesion within a patient.

Chapter 5

98

Figure 2. Representative [18F]FDOPA PET scan.[18F]FDOPA uptake per lesion in patient 15. A) [18F]FDOPA PET scan B) [18F]FDOPA PET scan fused with low dose computed tomographic scan. Written informed consent for publication of the clinical images was obtained from the patient.

Figure 3. [18F]FDOPA uptake per tumor lesion according to location in the body. SUVpeak per tumorVOI according to location. Each symbol represents a single lesion. Kind of symbol is equal per lesion in the same kind of organ.LN= lymph node, SUV=standard uptake value, VOI=volume of interest.

5

99


Figure 4. SUVmean per VOI according to location in the body of non-tumoral tissue.The VOIs per organ were of the same size per organ. The bar is showing the median. Each symbol represents a single VOI. Kind of symbol is equal per VOI in the same kind of organ.SUV=standard uptake value, VOI=volume of interest.

Table 3. Median SUV mean value of each background organ and correlation with tissue den-sity. Organ SUV mean median

(range)Correlation SUV mean with tissue density (Pearson’s r)

p-value

Kidney 2.8 (2.3 -3.3) 0.04 NS

Liver 1.7 (1.3-1.9) 0.01 NS

Pancreas 1.5 (1.2-1.7) -0.09 NS

Myocardium 1.4 (1.2-1.6) -0.11 NS

Spleen 1.2 (1.0-1.4) -0.16 NS

Muscle 1.2 (1.1-1.3) -0.09 NS

Bone marrow 1.1 (0.8-1.3) 0.10 NS

Bloodpool 1.1 (0.9-1.2) 0.00 NS

Brain 1.1 (0.8-1.3) 0.07 NS

Cortical bone 0.9 (0.7-1.1) -0.17 NS

Lung 0.2 (0.2-0.3) 0.83 1.28 *10-10

Total 1.2 (0.9-1.5) 0.22 5*10-6

NS = not significant, SUV=standard uptake value.

Chapter 5

100

Figure 5. Partial volume effect compared with a phantom model.SUV peak of the lesion versus the volume of the lesion, compared with the curve of the ‘partial volume effect’ measured with a phantom. cc= cubic centimeters, SUV=standard uptake value, VOI=volume of interest.

Figure 6. SUVmean per VOI according to tissue density.SUV=standard uptake value, VOI=volume of interest.

5

101


Discussion

[18F]FDOPA PET scans showed interlesional tumor heterogeneity within patients with SI-NET and identified lesions with mild or extreme high [18F]FDOPA uptake in 60% of the evaluable patients. Interlesional tumor heterogeneity of NET tumor characteristics within patients has been shown in several pathology studies (4-6,19). Currently no stud-ies aiming to identify lesions with high tracer uptake are performed. We performed a refined analyses of tumor burden with the use of tumor to background ratio to compen-sate for differences between organs in tracer uptake, vessel density and permeability (17,20). With the knowledge that interlesional tumor heterogeneity in neuroendocrine tumors can be determined with [18F]FDOPA PET, we were interested if lesions with mild or extreme high [18F]FDOPA uptake could be identified. Because total tumor burden measured on [18F]FDOPA PET, correlate with 5-HIAA in urine, a better insight in heterogeneity could potentially be used in case of systemic symptoms (10). In patients with symptomatic liver metastases, local treatment with RFA, TACE, TAE, resection or SIRT are relevant treatment modalities (9). With the selection of tumor lesions with a relatively high tumor burden, local treatments might be more effective in symptom management, than selection based on other characteristics. In general, tumor heterogeneity in neoplasms is determined by genetic variation be-tween subclones of tumor cells. As a consequence cell clones that have an advantage within a given tumor micro-environmental context survive and can expand through Darwinian selection (2). This may lead to tumor adaptation and therapeutic failure (21). Interlesional tumor heterogeneity on [18F]FDOPA PET was rudimentary analyzed in an imaging study in 77 patients with NET. The SUV max, without correction for the SUV of the background organ, between lesions within the patient, varied up to 29-fold (10). In NET one study analyzed heterogeneity within a single tumor lesion. 141 patients with NET, of which 108 patients had a GEP-NET, underwent a Gallium-68 somatostatin ana-logue ([68Ga]SSA) PET scan to select them for radiolabeled peptide receptor therapy. Heterogeneity within a lesion was shown to be present and entropy (a heterogeneity parameter) appeared to be predictive for progression free survival and overall survival (22). Recently, it was shown that total tumor burden in patients with SI-NET measured with [68Ga]SSA PET scan correlated with tumor markers; with 5-HIAA level in 24h urine and with chromogranine A, however this study did not investigate heterogeneity between tumor lesions (23). In another study, in five patients the genetic heterogeneity between primary and metastatic lesions was evaluated by whole exome sequencing. This revealed a highly varying degree of genetic heterogeneity between primary lesions and hepatic metastasis (24). In studies where sensitive molecular scans are used like in this study, many small tumor lesions could become visible. In small lesions, the partial volume effect (PVE) has to

Chapter 5

102

be taken into account. Part of variability in SUV, might be explained by the PVE. This is demonstrated in the non-linear curve of our data, similar to the curve derived from the phantom model (Fig 5) (25,26). In total 90% of lesions in our study were smaller than 20 mL and therefore their SUVs could be influenced by partial volume effects (PVE). A component of PVE is tissue fraction effect (27). Healthy lung tissue has a lower den-sity because it contains air. It is found that an increase in observed SUV is associated with regions of increased lung density (27,28). In our analysis this is reflected in a linear correlation between lung density with the SUV and a lower SUV mean of lung tissue compared to SUV mean value of the other organs used for background measurement (Fig 6). This resulted in higher [18F]FDOPA uptake in lung lesions compared with lesions in other organs (Fig 3). Several methods are developed to correct for tissue fraction effect in lung (27,28). In follow-up studies, to select lesions with substantially higher [18F]FDOPA uptake for local treatment, the tissue fraction effect and PVE have to be taken into account.

ConclusionsIn patients with SI-NET, [18F]FDOPA uptake shows considerable heterogeneity in uptake between tumor lesions within a patient. Furthermore in some patients lesions with mild or extreme high [18F]FDOPA uptake could be identified. This suggests that [18F]FDOPA PET scans can serve, in patients with systemic symptoms, due to an overproduction of catecholamines, to select lesions with substantially higher [18F]FDOPA uptake for local treatment.

5

103


References 1. Raut CP, Kulke MH, Glickman JN, Swanson RS, Ashley SW. Carcinoid tumors. Curr Probl

Surg 2006;43:383-450. 2. Burrell RA, McGranahan N, Bartek J, Swanton C. The causes and consequences of genetic

heterogeneity in cancer evolution. Nature 2013;501:338-45. 3. Couvelard A, Deschamps L, Ravaud P, et al. Heterogeneity of tumor prognostic markers:

a reproducibility study applied to liver metastases of pancreatic endocrine tumors. Mod Pathol 2009;22:273-81.

4. Shi C, Gonzalez RS, Zhao Z, et al. Liver metastases of small intestine neuroendocrine tumors: Ki-67 heterogeneity and World Health Organization grade discordance with primary tumors. Am J Clin Pathol 2015;143:398-404.

5. Richards-Taylor S, Tilley C, Jaynes E, et al. Clinically significant differences in Ki-67 pro-liferation index between primary and metastases in resected pancreatic neuroendocrine tumors. Pancreas 2017;46:1354-8.

6. Charoenpitakchai M, Liu E, Zhao Z, et al. In liver metastases from small intestinal neuroen-docrine tumors, SSTR2A expression is heterogeneous. Virchows Arch 2017;470:545-52.

7. Pavel M, Baudin E, Couvelard A, et al. ENETS Consensus Guidelines for the manage-ment of patients with liver and other distant metastases from neuroendocrine neoplasms of foregut, midgut, hindgut, and unknown primary. Neuroendocrinology 2012;95:157-76.

8. Kennedy A, Bester L, Salem R, et al. Role of hepatic intra-arterial therapies in metastatic neuroendocrine tumours (NET): guidelines from the NET-Liver-Metastases Consensus Conference. HPB (Oxford) 2015;17:29-37.

9. van Hazel GA, Heinemann V, Sharma NK, et al. SIRFLOX: Randomized phase III trial comparing first-line mFOLFOX6 (plus or minus bevacizumab) versus mFOLFOX6 (plus or minus bevacizumab) plus selective internal radiation therapy in patients with metastatic colorectal cancer. J Clin Oncol 2016;34:1723-31.


11. Kema IP, de Vries EG, Schellings AM, Postmus PE, Muskiet FA. Improved diagnosis of carcinoid tumors by measurement of platelet serotonin. Clin Chem 1992;38:534-40.

12. Luurtsema G, Boersma HH, Schepers M, et al. Improved GMP-compliant multi-dose pro-duction and quality control of 6-[18F]fluoro-L-DOPA. EJNMMI Radiopharm Chem 2017;1:7.

13. Koopmans KP, de Vries EG, Kema IP, et al. Staging of carcinoid tumours with 18F-DOPA PET: a prospective, diagnostic accuracy study. Lancet Oncol 2006;7:728-34.

14. Hudson HM, Larkin RS. Accelerated image reconstruction using ordered subsets of projec-tion data. IEEE Trans Med Imaging 1994;13:601-9.

15. van Asselt SJ, Oosting SF, Brouwers AH, et al. Everolimus reduces 89Zr-bevacizumab tumor uptake in patients with neuroendocrine tumors. J Nucl Med 2014;55:1087-92.

16. Koopmans KP, Neels OC, Kema IP, et al. Improved staging of patients with carcinoid and islet cell tumors with 18F-dihydroxy-phenyl-alanine and 11C-5-hydroxy-tryptophan positron emission tomography. J Clin Oncol 2008;26:1489-95.

17. Frings V, van Velden FH, Velasquez LM, et al. Repeatability of metabolically active tumor volume measurements with FDG PET/CT in advanced gastrointestinal malignancies: a multicenter study. Radiology 2014;273:539-48.

Chapter 5

104

18. National Electrical Manufacturers Association. NEMA Standards publication NU 2-2007, Performance measurements of positron emission tomographs. 2007.

19. Francis JM, Kiezun A, Ramos AH, et al. Somatic mutation of CDKN1B in small intestine neuroendocrine tumors. Nat Genet 2013;45:1483-6.

20. Bol K, Haeck JC, Groen HC, et al. Can DCE-MRI explain the heterogeneity in radiopep-tide uptake imaged by SPECT in a pancreatic neuroendocrine tumor model? PLoS One 2013;8:e77076.

21. Gerlinger M, Rowan AJ, Horswell S, et al. Intratumor heterogeneity and branched evolution revealed by multiregion sequencing. N Engl J Med 2012;366:883-92.

22. Werner RA, Lapa C, Ilhan H, et al. Survival prediction in patients undergoing radionu-clide therapy based on intratumoral somatostatin-receptor heterogeneity. Oncotarget 2017;8:7039-49.

23. Tirosh A, Papadakis GZ, Millo C, et al. Association between neuroendocrine tumors bio-markers and primary tumor site and disease type based on total (68)Ga-DOTATATE-Avid tumor volume measurements. Eur J Endocrinol 2017;176:575-82.

24. Walter D, Harter PN, Battke F, et al. Genetic heterogeneity of primary lesion and metastasis in small intestine neuroendocrine tumors. Sci Rep 2018;8:3811.

25. Hoetjes NJ, van Velden FH, Hoekstra OS, et al. Partial volume correction strategies for quantitative FDG PET in oncology. Eur J Nucl Med Mol Imaging 2010;37:1679-87.

26. Hoffman EJ, Huang SC, Phelps ME. Quantitation in positron emission computed tomogra-phy: 1. Effect of object size. J Comput Assist Tomogr 1979;3:299-308.

27. Holman BF, Cuplov V, Millner L, et al. Improved correction for the tissue fraction effect in lung PET/CT imaging. Phys Med Biol 2015;60:7387-402.

28. Lambrou T, Groves AM, Erlandsson K, et al. The importance of correction for tissue fraction effects in lung PET: preliminary findings. Eur J Nucl Med Mol Imaging 2011;38:2238-46.

5

105


All 18F-DOPA PET/CT scans made in adults in the UMCG from 4-2-2014 to 1-4-2015.

(Excluding PET only camera) n=182 scans

(n=160 patients)

Only patients with SI-NETn=61 scans

(n=48 patients)

Only patients wtih ≥ 1 positive lesion on 18F-DOPA PET/scan

n=93 scans (n=77 patients)

Exclusion ≤ 1 positive lesion on 18F-DOPA PET/scan: n=89 scans

(n= 83 patients)

Only patients with CT and 18F-DOPA PET scan

n=48 scans -> (n=38 patients)

Exclusion patients with primary tumor of pancreas or bronchopulmonal

n= 32 scans (n=29 patients)

Exclusion no CT-scan was made within 6 months before or after the 18F-DOPA PET/scan

n=13 scans (n=10 patients)

Figure S1. Diagram of in- and exclusion

Chapter 5

106

Organ

Num

ber

of le

sion

s

BoneBowel

Heart

Liver

Lung

LN/mes

enter

ium

Pancre

asOther

0

100

200

300

400

500

PET lesionsCT inconclusive or malignantlesions

Organ

Nr o

f les

ions

BoneBowel

Heart

Liver

Lung

LN / mes

enter

ium

Pancre

asOther

0

100

200

300

400

500

PET lesions Malignant CT lesions

Figure S2. [18F]FDOPA PET lesions versus inconclusive and/or malignant lesions on CT.

5

107


Patients

18F-

DOPA

* vo

lum

e (m

L)

1 2 3 4 5 6 7 8 9 10111213141516171819202122232425262728293031323334353637380

1000

2000

3000

4000

5000

Figure S3. Tracer uptake multiplied by volume per lesion per patient. Each symbol represents a single lesion. Color and kind of symbol are equal per lesion within a patient.

6

Reasons for the cold immune microenvironment of

neuroendocrine tumors

L.D. de Hosson1‡, G. Bouma1‡, G. Kats-Ugurlu2, M. Bulthuis2, E.G.E. de Vries1, M. van Faassen3, I.P. Kema3, A.M.E. Walenkamp1

‡ The first two authors contributed equally to the manuscript

1Department of Medical Oncology, University of Groningen, University Medical Center Groningen, Groningen, The Netherlands

2Department of Pathology, University of Groningen, University Medical Center Groningen, Groningen, The Netherlands

3Department of Laboratory Medicine, University of Groningen, University Medical Center Groningen, Groningen, The Netherlands

Chapter 6

110

Abstract

Tumors can escape the immune system by expressing programmed death-ligand-1 (PD-L1) which allows them to bind to PD-1 on T-cells and avoid recognition by the im-mune system. In addition, T-cells, indoleamine 2,3-dioxygenase (IDO) and tryptophan 2,3-dioxygenase (TDO) play a role in immune suppression. There is limited knowledge of interaction of neuroendocrine tumors (NETs) with their immune microenvironment and the role of immunotherapy in patients with NET. This study aimed to investigate the immune microenvironment of serotonin-producing (SP) and non-serotonin-producing NETs (NSP-NETs).Tumors of 33 patients with SP-NET and 18 patients with NSP-NETs, were studied. Im-munohistochemically analyses were performed for PD-L1, T-cells, IDO, TDO, mismatch repair proteins (MMRp) and activated fibroblasts. PD-L1 expression was seen in <1% of tumor and T-cells. T-cells, were present in 33% of NETs, varying between 1-10% T-cells per high power field. IDO was expressed in tumor cells in 55% of SP-NETs and 22% of NSP-NETs (p=0.039). TDO was expressed in stromal cells in 64% of SP-NETs and 13% NSP-NETs (p=0.001). None of the tumors had loss of MMRp. TDO expressing stromal cells also strongly expressed α-SMA and were identified as cancer associated fibroblasts (CAFs). In conclusion, NETs lack pre-existing immunity as they do not express PD-L1, contain only few T-cells and were not MMRp deficient. The expression of IDO and TDO in a sub-stantial part of NETs and the presence of CAFs suggest two mechanisms responsible for the cold immune microenvironment. These mechanisms can be further explored to enhance anti-tumor immunity and clinical responses.

6

111

Reasons for the cold immune microenvironment of neuroendocrine tumors.

Introduction

Neuroendocrine tumors (NETs) comprise a heterogeneous group of tumors, which are predominantly derived from the enterochromaffin cells of the gastro-enteropancreatic tract or bronchopulmonary system (1). NETs can produce various biogenic amines and polypeptide hormones of which serotonin is the most common (2,3). Radical resection of the NET is the only possibility to cure. However, patients with NET often present with non-resectable or metastatic disease. Non-curative systemic treatment aimed at controlling symptoms and progression of disease includes somatostatin analogues, in-terferon, everolimus, sunitinib, peptide receptor radionuclide therapy and chemotherapy (4). None of the currently used systemic treatments for GEP-NET could be graded as substantial clinical beneficial according to the ESMO-magnitude of clinical benefit scale (5). Therefore, there is an unmet need for new and better systemic treatment modalities. Over the last years, immunotherapy with immune checkpoint inhibiting antibodies targeting the cell surface proteins programmed death-ligand-1 (PD-L1) on tumor and (non-)immune cells and programmed death-1 (PD-1) on mono-/lymphocytes have shown antitumor activity across numerous tumor types (6,7). Targeting of PD-L1 and PD-1 lead to activated T-cells in the tumor microenvironment. There is however limited information as yet with regards to the activity of these drugs in NETs. There are a number of other factors considered to be associated with a response to checkpoint inhibitor treatment, such as presence of T-cells, and high mutational tumor load (8,9). In addition, there is a major interest in the tryptophan-degrading enzymes indoleamine 2,3-dioxygenase (IDO) and tryptophan 2,3-dioxygenase (TDO) along the kynurenine pathway that depletes tryptophan in the tumor microenvironment (10-12). IDO and TDO are especially of interest in NETs as these tumors often produce serotonin which potentially depletes its precursor, tryptophan (13).Overall, strikingly little is known about the complex interactions of NET cells with their surrounding immune microenvironment. Consequently knowledge about potential tar-gets for immunotherapy in patients with NET is limited (14-16). Therefore, the aim of this study was to investigate the tumor immune microenvironment, i.e. the presence of PD-L1, T-cells, IDO, TDO, MMRp and CAFs in tissue of treatment-naive SP-NET and NSP-NET patients.

Material and methods

ParticipantsMedical records of newly referred NET patients to the Department of Medical Oncology of the University Medical Center Groningen (UMCG) between January 1, 2008 and

Chapter 6

112

December 31, 2014 were screened. Patients diagnosed with a NET grade 1 or 2 ac-cording to the World Health Organization 2010 classification were selected. A patient was diagnosed with a SP-NET when urinary 24-h excretion of 5-HIAA was >3.8 mmol/mol creatinine and/or platelet serotonin >5.4 nmol/109 platelets (12). Serotonin produc-tion and 5-hydroxyindolacetic acid (5-HIAA) in 24 hour (24-h) urine were measured by high performance liquid chromatography.(17) Included were NET patients with tumor tissue, platelet rich plasma for analysis of se-rotonin and/or urinary 24-h excretion of 5-hydroxyindolacetic acid (5-HIAA) available before start of systemic antitumor treatment. Somatostatin analogue use was allowed for maximal 14 days before tumor tissue was collected. Excluded were NET patients with other primary solid/haematological malignancy, auto-immune disease (e.g. colitis) or infectious disease (e.g. hepatitis) as well as patients with concomitant use of treat-ment interfering with IDO-activity (e.g. interferon).Patients were clinically staged according to the Union for International Cancer Control (UICC) guidelines (18). Histopathological analysis of the formalin-fixed paraffin-embed-ded tumor tissue of the patients was centrally reviewed by a dedicated NET pathologist (GKU).Since residual archival material was studied which does not interfere with patient care and does not involve the physical involvement of the patient, no ethical approval is required for this study according to Dutch legislation (the Medical Research Involving Human Subjects Act) (19). Despite this, all patients alive (n=20) were approached and gave written informed consent to use their residual material.

Tumor histology and ImmunohistochemistryIn all tumor samples 3 µm slides of FFPE tumor samples were studied for morphology and mitotic count on standard hematoxylin and eosin (H&E) stain and proliferation index was determined using immunohistochemistry based Ki-67 stain. (mouse anti-Ki67, MiB-1 clone, dilution 1:100 Dako, Glostrup, Denmark). Two antibodies for PD-L1 staining were used, namely mouse anti-PD-L1 (clone 22C3, 1:50, Dako, Glostrup, Denmark) and rabbit anti-PD-L1 (clone E1L3N, 1:200 Cell Signaling Technology, Danvers, MA, USA). PD-L1 antibodies were applied in the Ventana Ultra staining system. To detect MMR antigens, anti-MLH-1 mouse monoclonal primary antibody (clone M1, Roche Diagnos-tics, IN, USA), PMS2 rabbit monoclonal antibody (clone EPR3947, Cell Marque, CA, USA), MSH2 mouse monoclonal antibody (clone G219-1129, Cell Marque), CONFIRM anti-MSH6 mouse monoclonal primary antibody (clone 44, Roche Diagnostics) were used. The mouse-anti-CD3 antibody (1:50, Monosan, Sanbio, Uden, the Netherlands) served to recognize T-cells. Antibodies against IDO (mouse anti-IDO, MAB5412, 1:25, Chemicon, Millipore, Amsterdam, the Netherlands), TDO (rabbit anti-TDO2, clone HPA 039611, 1:200, Atlas Antibodies, Bromma, Sweden) and α-SMA (mouse anti-SMA,

6

113


clone 1A4, 1:1000, Sigma, MI, USA) were applied. α-SMA is a well-established marker for myofibroblasts and myofibroblast-like cells in the tumor microenvironment, also known as CAFs (20). One pathologist (GKU) and two researchers (GB, LDH) evaluated the stained slides at a double-head microscope blinded for clinical and histopathological information. To avoid learning effect, every first 10 slides that were scored and slides that were dif-ficult to score, were scored again without knowledge of the first score. Positive external controls were placenta for PD-L1, appendix for the MMR antigens and CD3, lymph node for IDO and prostate for TDO. For α-SMA the pericytes of blood vessels served as internal control.The tumor and its immediate environment were evaluated. Tumor cells with ≥ 1% positive staining for PD-L1, IDO and TDO were considered as positive. MMRp loss was considered when tumor cell nucleus showed no staining in comparison to internal inflammatory cells or appendix as external control. T-cells were scored ‘present’ if ≥1% of the cells in a high power field composed of T-cells distributed patchy or diffusely in CD3. PD-L1 showed a membranous staining pattern of tumor cells, IDO showed a cytoplasmic staining pattern with presence of acellular small depositions. TDO showed besides a cytoplasmic staining pattern more remarkable expression in tumor stroma. Finally, a α-SMA staining was performed to further characterise TDO positive stromal cells, on five tumors with TDO expression in stroma, and five without TDO expression in stroma.

Statistical analysisFor this exploratory study no sample size calculation was performed. Descriptive statistics (e.g. median, ranges, and frequencies) were calculated for all measures. Mann-Whitney U test was used to compare distributions across groups. Associations of categorical variables were tested using the Chi-Square test. Tests were performed two-sided, and P values <0.05 were considered significant. Analyses were executed using the software package SPSS, version 23 for Windows (SPSS, Inc, Chicago, IL, USA).

Results

Patient characteristics Clinical and pathological characteristics at the moment of tumor collection of all SP-NET (N=33) and NSP-NET (N=18) patients are summarized in Table 1. Thirteen SP-NET pa-tients were shortly treated with somatostatin analogues before the tissue sampling. The short use of somatostatin analogues was unavoidable in 13 out of 51 (25%) patients since this was prescribed to prevent a carcinoid crisis during an invasive procedure.

Chapter 6

114

Expression of PD-L1, MMR, presence of tumor infiltrating T-cells, IDO and TDO in NETs.None of the tumors or T-cells were positive for PD-L1 in anti-22C3 or anti-E1L3N stains. None of the tumors showed loss of the MMRp, T-cells were present in 15 of the 45 samples, varying between 1-10% T-cells per HPF (Table 2). T-cells were most frequently found within the stroma of NETs of the jejunum/ileum which were all SP-NETs (Table 3). IDO expression was restricted to tumor cells and varied between focal and diffuse pres-ence of intracytoplasmic acellular small depositions. IDO expression in tumor cells was

Table 1. Clinicopathological characteristics of patients. NET patients

(N=51) ANSP NET

(N=18)SP NET(N=33)

Mean age in years ± SD 62.8 ±11.0

Male gender 30 (59) 11 (61) 19 (58)

Primary tumor location Lung Stomach Duodenum Pancreas Jejunum/ileum Colon/rectum Unknown

2 (4)1 (2)2 (4)

14 (27)22 (43)2 (4)8 (16)

2 (11)1 (6)

2 (11)11 (61)0 (0)1 (6)1 (6)

0 (0)0 (0)0 (0)3 (9)

22 (67)1 (3)

7 (21)

Tumor grade Grade 1 Grade 2 Unknown

34 (67) 16 (31) 1 (2)

8 (44)10 (56)0 (0)

25 (76)6 (18)2 (6)

Disease stage Stage 1/2 Stage 3/4 Unknown

3 (6)46 (90)2 (4)

2 (11)15 (83)1 (6)

1 (3)31 (94)1 (3)

Source tissue sample Primary tumor Metastasis Unknown

30 (59)20 (39)1 (2)

10 (56)8 (44)0 (0)

20 (61)12 (36)1 (3)

Location tissue sample Liver Lymph node Jejunum/ileum Pancreas Duodenum Lung OtherB

14 (27)4 (8)

18 (35) 5 (10)3 (6)2 (4)

5 (10)

8 (44)1 (6)0 (0)

3 (17)3 (17)1 (6)2 (6)

6 (18)3 (9)

18 (55)2 (6)0 (0)1 (3)3 (9)

A Values are reported as number (percentage) unless noted otherwiseB Other sites of the tissue sample collection were mesenterium of the small intestine (N=2), colon (N=1), stomach (N=1), peritoneum (N=1).NET; neuroendocrine tumor, NSP-NET; non-serotonin-producing neuroendocrine tumor, SP-NET; serotonin-pro-ducing neuroendocrine tumor.

6

115


more frequently observed in SP-NETs (55%, 18/33) as compared to NSP-NETs (22%, 4/18) (p=0.026, Table 2).

There were three different patterns of TDO expression in the NETs, namely; in tumor cells, in stroma or in both (Figure 1B and 1C). The majority of NETs expressed TDO either in the tumor cells (37%, 17/46) or stroma (44%, 18/41) (Table 2). Remarkably, TDO in stroma was observed in 64% (16/25) evaluable SP-NETs and 13% (2/16) of the NSP-NETs. To investigate the origin of these TDO positive stromal cells, a α-SMA staining was performed on 10 slides. Tumors with TDO expression in stromal cells strongly expressed α-SMA, and were therefore identified as CAFs.

Table 2. Presence of T-cells, expression of IDO, TDO, PD-L1 and MMRp in NETs. All NETn (%)

NSP-NET n (%)

SP-NETn (%)

p-value

Tissue samples 51 (100) 18 (100) 33 (100)

PD-L1 expression (22C3 antigen)NegativePositive

51 (100)0 (0)

18 (100)0 (0)

33 (100)0 (0)

NS

PD-L1 expression ( E1L3N antigen)NegativePositive

51 (100)0 (0)

18 (100)0 (0)

33 (100)0 (0)

NS

MMRp Loss of MMRp No loss of MMRp

0 (0)51 (100)

0 (0)18 (100)

0(0)33 (100)

NS

T-cellsAbsentPresentNot evaluable

30 (59)15 (29) 6 (12)

10 (56)6 (33)2 (11)

20 (61) 9 (27) 4 (12)

NS

IDO expressionNegativePositive

29 (57)22 (43)

14 (78) 4 (22)

15 (45)18 (55)

0.039

TDO expression in tumor cells NegativePositive Not evaluable

29 (63)17 (37)5 (10)

11 (65) 6 (35)1 (6)

18 (62) A

11 (38)4 (12)

NS

TDO expression in stroma NegativePositiveNot evaluable

23 (56)18 (44)10 (20)

14 (88)2 (13)2 (11)

9 (36)16 (64)8 (24)

0.001

A Three tissue samples of SP-NETs with not evaluable stroma were TDO positive in the tumor.IDO; indoleamine 2,3-dioxygenase, MMRp; mismatch repair proteins, NET; neuroendocrine tumor, NSP-NETs; non-serotonin-producing neuroendocrine tumor, PD-L1; programmed death-ligand 1, SP-NETs; serotonin-produc-ing neuroendocrine tumor, TDO; tryptophan 2,3-dioxygenase.

Chapter 6

116

Figure 1 Expression of IDO, TDO and α-SMA and IDO in a serotonin producing NET of the ileum.Illustrative images of indoleamine 2,3-dioxygenase (IDO) and tryptophan 2,3-dioxygenase (TDO) expression in a serotonin producing neuroendocrine tumor (NET) in ileum resection specimen. Diff erent compartments of ileum; #intestinal epithelium, ^submucosal tumor, *muscularis propria (A HE, 20x). Magnifi cation of the submucosal NET (B HE, 200x). In lower magnifi cation IDO is not detectable (C, IDO, 20x) In higher magnifi cation a diff use, strong brown intracytoplasmic, dot-like IDO expression (marked by arrows) is seen in tumor cells already in low magnifi ca-tion (D IDO, 1000x). TDO expression is visible in stromal cells surrounding the tumor cells (E TDO, 20x, F TDO, 200x). α-SMA is expressed between tumor cells and has a stronger expression in areas with more TDO expression (G α-SMA, 20x, H α-SMA, 200x).

6

117


Discussion

We explored the immune microenvironment of NETs. In NETs both tumor and T-cells did not express PD-L1. In addition, only the minority of NETs contained a limited number of T-cells. A substantial part of NETs expressed IDO and TDO. TDO was also expressed by stromal cells. None of the tumors were MMRp deficient. α-SMA staining of TDO expressing stromal cells identified these cells as CAFs (20). Our study showed that NETs exhibit a ‘cold’ tumor microenvironment lacking several characteristics which are currently associated with a response to checkpoint inhibitor treatment (6,7). The expression of IDO and TDO in a substantial part of NETs and the presence of CAFs suggest two mechanisms responsible for the cold immune microen-vironment. Interestingly, immunohistochemically analysis of lung cancer specimens revealed increased TDO2 expression in CAFs. Furthermore, in a mouse model, administration of TDO2 inhibitor improved T-cell response, and decreased tumor metastasis in mice with metastatic lung cancer (21,22). The enzymes IDO and TDO degrade tryptophan

Table 3. Presence of T-cells and immunohistochemical expression of IDO, TDO and PD-L1 classified by primary origin of the NET

Jejunum/ileum n

Pancreas n Lung n Unknown n Other na

Tissue samples 22 14 2 8 5

T-cellsAbsentPresentNot evaluable

1372

1112

011

530

131

IDO expressionNegativePositive

913

95

20

62

32

TDO expression in tumour cells NegativePositive Not evaluable

1543

662

200

350

320

TDO expression in stroma NegativePositiveNot evaluable

2155

1211

101

512

311

PD-L1 expression 0 0 0 0 0

MMRp 0 0 0 0 0A Other sites of primary origin of the NET are stomach (N=1), duodenum (N=2), colon (N=1), rectum (N=1)B One serotonin producing NET with not evaluable stroma was TDO positive in the tumorIDO; indoleamine 2,3-dioxygenase, MMRp; mismatch repair proteins, NET; neuroendocrine tumor, PD-L1; pro-grammed death-ligand 1, TDO; tryptophan 2,3-dioxygenase.

Chapter 6

118

into kynurenines resulting in a tryptophan depleted tumor microenvironment rich of kynurenines. This microenvironment leads to suppression of effector T-cells or the conversion to tumor tolerant regulatory T-cells (Tregs) (10-12). A number of IDO and TDO inhibitors advanced into clinical trials with or without PD-1 antibody inhibitors in patients with solid tumors. Also an IDO1/TDO dual inhibitor is currently tested in a phase I study (23-27). Surprisingly, the ECHO-301/KEYNOTE 252 study, was recently announced to be terminated early due to failure to improve PFS (28). In this phase III study patients with unresectable or metastatic melanoma were randomized to receive pembrolizumab in combination with either epacadostat (an IDO-inhibitor) or placebo. According to experts the IDO strategy has been moved to randomized clinical trials too fast (29). Our data show that especially serotonin producing NET tumors express IDO and TDO. These patients often have low tryptophan levels as tryptophan is the sole precursor of peripherally and centrally produced serotonin. Potentially this means that in patients with low tryptophan levels, IDO-mediated immune suppression is most prominent, and these patients might therefore be interesting candidates for combination treatment with immune checkpoint inhibitors combined with IDO inhibitors. In our study we could not find an association between IDO expression and presence of T-cells and this might be an indication of the complex interaction of tumors and their (immune) microenvironment. Preclinical studies in mouse models of various cancer types showed that CAFs together with other stromal cells are responsible for restricting the accumulation of T-cells in the vicinity of cancer cells (30). Furthermore, CAFs act by secretion of various growth factors like transforming growth factor-beta (TGF-β) (31,32). Increased TGF-β in the tumor immune microenvironment is recently recognised to represent a primary mechanism of immune evasion that promotes T-cell exclusion (33). In patients with metastatic urothelial carcinoma, lack of response to immune checkpoint blockade is associated with increased TGF-β signalling in fibroblasts in the tumor microenvironment (33). Combining TGF-β blockade with immune checkpoint blockade in mouse models increases the anti-tumor efficacy of the therapy, suggesting that identifying and target-ing microenvironmental regulators of anti-tumor immunity may increase the reach of immune-therapeutical approaches (34).Limited PD-L1 expression in NET cells has also been observed in six studies while in two other studies more than half of NETs had PD-L1 expression (Table S1) (35-42). The use of different PD-L1 antibodies across the studies may account for heterogeneous findings. Therefore we used, two validated PD-L1 antibodies, 22C3 and E1L3N, which both showed no expression of PD-L1 in NET cells and T-cells.Grade 3 neuroendocrine neoplasms (NENs) were in contrast to NETs characterized by strong PD-L1 expression (39). A correlation between tumor grade and PD-L1 expression was also seen in neuroendocrine neoplasms of pulmonary origin (42,43). One study

6

119


investigating 32 patients treated with systemic therapy for GEP-NENs found that PD-L1 expression was associated with progression free survival (PFS) (37). Furthermore a multivariate analysis of 80 pulmonary NENs revealed that PD-L1 expression, PD1 expression, and distant metastasis of pulmonary NENs were independently associated with survival time (42).In the current study a paucity of T-cells is seen, conform several other studies analyz-ing NET immunohistochemically (44,45). In one study samples from 31 midgut NETs (grade 1 and 2) were analyzed. Overall, the tumors contained a higher proportion of Tregs compared with matched normal tissue (45). Other studies examining T-cells in neuroendocrine neoplasms reported more CD3 expressing T-cells. In another study, 68% of pancreatic NETs (N=87) and 97% of NET liver metastases of patients (N=39) with various primary tumor sites contained >10 intratumoral T-cells per 10 HPF. In these two groups, Tregs were found in 34% and 33%, respectively (15). Others found tumor infiltrating lymphocytes (TILs) in 14 out of 17 samples, varying from single positive cells to multiple positive cells (16).The paucity of T-cells in NETs and the relatively high rate of Tregs, might be due to CAFs and their secreted factors as well as of IDO and TDO expression in the tumor and microenvironmental cells. In other tumor types, such as colorectal, oesophageal and endometrial cancer, an inverse correlation between IDO expression and T-cells has been observed (46-48).The PD-1 antibodies pembrolizumab and nivolumab are registered for the treatment of several tumor types. Furthermore nivolumab is registered for MMRp deficient meta-static colorectal cancer and pembrolizumab for MMRp deficient tumors, irrespective of the primary tumor (8,9,49). In MMRp-proficient tumors, probable DNA base pairing errors are corrected in newly replicated DNA, leading to microsatellite stable tumors. In our study all investigated NETs were MMRp-proficient. This is in line with a study of 35 pancreatic NETs and 34 small intestinal NETs, where all pNETs and 31 small intestinal NETs were MMRp-proficient and with another study where all included 70 pancreatic NETs were microsatellite stable (40,50).Scarce data exist about treatment with immune checkpoint inhibition in patients with NET. A small case series describes four patients with NET treated with checkpoint inhibitors. All patients showed improved quality of life and a PFS of >3 months, with two out of four still on therapy (51). In the multicohort phase 1B KEYNOTE-028 study, 41 pretreated patients with pancreas, lung and gut NET were treated with pembrolizumab. Results showed four patients with objective response and 29 patients with stable disease. Eight patients had severe (grade ≥3) treatment related adverse events (52). Multiple phase 1 and 2 trials are currently recruiting patients with NET for immune checkpoint inhibitor treatment (51).

Chapter 6

120

Our findings demonstrate that in patients with NET both tumor and CAFs express IDO and TDO. Therefore we hypothesise that patients with NET will potentially benefit from combination immunotherapies to overcome this cold immune microenvironment.

6

121


References 1. Yao JC, Hassan M, Phan A, et al. One hundred years after “carcinoid”: epidemiology of and

prognostic factors for neuroendocrine tumors in 35,825 cases in the United States. J Clin Oncol 2008;26:3063-72.

2. Oberg K. Neuroendocrine tumors of the digestive tract: impact of new classifications and new agents on therapeutic approaches. Curr Opin Oncol 2012;24:433-40.

3. Kema IP, de Vries EG, Slooff MJ, Biesma B, Muskiet FA. Serotonin, catecholamines, his-tamine, and their metabolites in urine, platelets, and tumor tissue of patients with carcinoid tumors. Clin Chem 1994;40:86-95.

4. Pavel M, Baudin E, Couvelard A, et al. ENETS Consensus Guidelines for the manage-ment of patients with liver and other distant metastases from neuroendocrine neoplasms of foregut, midgut, hindgut, and unknown primary. Neuroendocrinology 2012;95:157-76.

5. de Hosson LD, van Veenendaal LM, Schuller Y, et al. Clinical benefit of systemic treatment in patients with advanced pancreatic and gastrointestinal neuroendocrine tumours accord-ing to ESMO-MCBS and ASCO framework. Ann Oncol 2017;28:3022-7.

6. Chen DS, Mellman I. Elements of cancer immunity and the cancer-immune set point. Nature 2017;541:321-30.

7. Brody R, Zhang Y, Ballas M, et al. PD-L1 expression in advanced NSCLC: Insights into risk stratification and treatment selection from a systematic literature review. Lung Cancer 2017;112:200-15.

8. Le DT, Uram JN, Wang H, et al. PD-1 blockade in tumors with mismatch-repair deficiency. N Engl J Med 2015;372:2509-20.

9. Le DT, Durham JN, Smith KN, et al. Mismatch repair deficiency predicts response of solid tumors to PD-1 blockade. Science 2017;357:409-13.

10. Weinlich G, Murr C, Richardsen L, Winkler C, Fuchs D. Decreased serum tryptophan concentration predicts poor prognosis in malignant melanoma patients. Dermatology 2007;214:8-14.

11. Schrocksnadel K, Wirleitner B, Winkler C, Fuchs D. Monitoring tryptophan metabolism in chronic immune activation. Clin Chim Acta 2006;364:82-90.

12. Kema IP, Schellings AM, Hoppenbrouwers CJ, Rutgers HM, de Vries EG, Muskiet FA. High performance liquid chromatographic profiling of tryptophan and related indoles in body fluids and tissues of carcinoid patients. Clin Chim Acta 1993;221:143-58.

13. Bouma G, van Faassen M, Kats-Ugurlu G, de Vries EG, Kema IP, Walenkamp AM. Niacin (vitamin B3) supplementation in patients with serotonin-producing neuroendocrine tumor. Neuroendocrinology 2016;103:489-94.

14. Ryschich E, Autschbach F, Eisold S, Klar E, Buchler MW, Schmidt J. Expression of HLA class I/II antigens and T cell immune response in human neuroendocrine tumors of the pancreas. Tissue Antigens 2003;62:48-54.

15. Katz SC, Donkor C, Glasgow K, et al. T cell infiltrate and outcome following resection of intermediate-grade primary neuroendocrine tumours and liver metastases. HPB (Oxford) 2010;12:674-83.

16. Vikman S, Sommaggio R, De La Torre M, et al. Midgut carcinoid patients display increased numbers of regulatory T cells in peripheral blood with infiltration into tumor tissue. Acta Oncol 2009;48:391-400.

Chapter 6

122

17. van de Merbel NC, Hendriks G, Imbos R, Tuunainen J, Rouru J, Nikkanen H. Quantitative determination of free and total dopamine in human plasma by LC-MS/MS: the importance of sample preparation. Bioanalysis 2011;3:1949-61.

18. Sobin LH, Compton CC. TNM seventh edition: what‘s new, what‘s changed: communication from the International Union Against Cancer and the American Joint Committee on Cancer. Cancer 2010;116:5336-9.

19. Central Committee on Research involving Human Subjects (Centrale Commissie Mensge-bonden Onderzoek). (2017, at http://www.ccmo.nl).

20. Pure E, Lo A. Can targeting stroma pave the way to enhanced antitumor immunity and immunotherapy of solid tumors? Cancer Immunol Res 2016;4:269-78.

21. Hsu YL, Hung JY, Chiang SY, et al. Lung cancer-derived galectin-1 contributes to cancer associated fibroblast-mediated cancer progression and immune suppression through TDO2/kynurenine axis. Oncotarget 2016;7:27584-98.

22. Yu CP, Song YL, Zhu ZM, Huang B, Xiao YQ, Luo DY. Targeting TDO in cancer immuno-therapy. Med Oncol 2017;34:73.

23. Pilotte L, Larrieu P, Stroobant V, et al. Reversal of tumoral immune resistance by inhibition of tryptophan 2,3-dioxygenase. Proc Natl Acad Sci U S A 2012;109:2497-502.

24. Mellemgaard A, Engel-Norregaard L, Andersen MH, Zocca M-, Svane IM. Combination immunotherapy with IDO vaccine and PD-1 inhibitors in advanced NSCLC. J Clin Oncol 2017;35.

25. Brochez L, Chevolet I, Kruse V. The rationale of indoleamine 2,3-dioxygenase inhibition for cancer therapy. Eur J Cancer 2017;76:167-82.

26. ClinicalTrials.gov NCT02073123 (2016, at https://clinicaltrials.gov/ct2/show/NCT02073123?term=ido&rank=16).

27. Prendergast GC, Malachowski WP, DuHadaway JB, Muller AJ. Discovery of IDO1 inhibi-tors: From bench to bedside. Cancer Res 2017;77:6795-811.

28. Incyte and Merck provide update on phase 3 study of epacadostat in combination with KEYTRUDA® (pembrolizumab) in patients with unresectable or metastatic melanoma. Business Wire. (2018, at https://www.businesswire.com/news/home/20180406005141/en/Incyte-Merck-Provide-Update-Phase-3-Study).

29. Garber K. A promising new cancer drug has hit a major setback, raising questions about whether the field is moving too fast. Science Magazine 2018; doi:10.1126/science.aau1287.

30. Joyce JA, Fearon DT. T cell exclusion, immune privilege, and the tumor microenvironment. Science 2015;348:74-80.

31. Kakarla S, Song XT, Gottschalk S. Cancer-associated fibroblasts as targets for immuno-therapy. Immunotherapy 2012;4:1129-38.

32. Kalluri R. The biology and function of fibroblasts in cancer. Nat Rev Cancer 2016;16:582-98.

33. Mariathasan S, Turley SJ, Nickles D, et al. TGFbeta attenuates tumour response to PD-L1 blockade by contributing to exclusion of T cells. Nature 2018;554:544-8.

34. Tauriello DVF, Palomo-Ponce S, Stork D, et al. TGFbeta drives immune evasion in geneti-cally reconstituted colon cancer metastasis. Nature 2018;554:538-43.

35. Kasajima A, Ishikawa Y, Iwata A, et al. Inflammation and PD-L1 expression in pulmonary neuroendocrine tumors. Endocr Relat Cancer 2018;25:339-50.

36. Lamarca A, Nonaka D, Breitwieser W, et al. PD-L1 expression and presence of TILs in small intestinal neuroendocrine tumours. Oncotarget 2018;9:14922-38.

6

123


37. Kim ST, Ha SY, Lee S, et al. The impact of PD-L1 expression in patients with metastatic GEP-NETs. J Cancer 2016;7:484-9.

38. da Silva A, Qian Z, Zhang S, et al. Immune checkpoint markers and immune response in well differentiated neuroendocrine tumors (NET) of the small intestine and pancreas NANETS 2016;:Abstract B3.

39. Cavalcanti E, Armentano R, Valentini AM, Chieppa M, Caruso ML. Role of PD-L1 expression as a biomarker for GEP neuroendocrine neoplasm grading. Cell Death Dis 2017;8:e3004.

40. Salem ME, Puccini A, Grothey A, et al. Landscape of tumor mutation load, mismatch repair deficiency, and PD-L1 expression in a large patient cohort of gastrointestinal cancers. Molecular Cancer Research 2018;16:805-812.

41. Cives M, Strosberg J, Coppola D. Pd1 and pdl1 expression in midgut neuroendocrine tumors. Neuroendocrinology 2016;103:36-7.

42. Fan Y, Ma K, Wang C, et al. Prognostic value of PD-L1 and PD-1 expression in pulmonary neuroendocrine tumors. Onco Targets Ther 2016;9:6075-82.

43. Kasajima A, Ishikawa Y, Iwata A, et al. PD-L1 Expression and its clinical relevance in neu-roendocrine tumors of the lung. ENETS 2018;:Abstract 2198.

44. Milione M, Pellegrinelli A, Centonze G, et al. T-cell infiltrate in GEP-NENS: Identifica-tion and evaluation of its distribution among the who G1, G2 and G3 NENS. Lab Invest 2016;96:151A.

45. Vesely C, Childs A, Wong YNS, Ogunbiyi O, Gander A, Luong TV. Systematic evaluation of the immune microenvironment of neuroendocrine tumors (NET). ENETS 2018;:Abstract 2199.

46. Brandacher G, Perathoner A, Ladurner R, et al. Prognostic value of indoleamine 2,3-dioxy-genase expression in colorectal cancer: effect on tumor-infiltrating T cells. Clin Cancer Res 2006;12:1144-51.

47. Ino K, Yamamoto E, Shibata K, et al. Inverse correlation between tumoral indoleamine 2,3-dioxygenase expression and tumor-infiltrating lymphocytes in endometrial cancer: its association with disease progression and survival. Clin Cancer Res 2008;14:2310-7.

48. Zhang G, Liu WL, Zhang L, et al. Involvement of indoleamine 2,3-dioxygenase in impairing tumor-infiltrating CD8 T-cell functions in esophageal squamous cell carcinoma. Clin Dev Immunol 2011;2011:384726.

49. Overman MJ, McDermott R, Leach JL, et al. Nivolumab in patients with metastatic DNA mismatch repair-deficient or microsatellite instability-high colorectal cancer (CheckMate 142): an open-label, multicentre, phase 2 study. Lancet Oncol 2017;18:1182-91.

50. Arnason T, Sapp HL, Rayson D, et al. Loss of expression of DNA mismatch repair proteins is rare in pancreatic and small intestinal neuroendocrine tumors. Arch Pathol Lab Med 2011;135:1539-44.

51. Chauhan A, Horn M, Magee G, et al. Immune checkpoint inhibitors in neuroendocrine tumors: A single institution experience with review of literature. Oncotarget 2017;9:8801-9.

52. Mehnert JM, Rugo HS, O’Neil BH, et al. Pembrolizumab for patients with PD-L1-positive advanced carcinoid or pancreatic neuroendocrine tumors: Results from the KEYNOTE-028 study. Annals of Oncology 2017;28:v142.

Chapter 6

124

Supplementary Figure S1. Overview of tryptophan metabolism; the kynurenine pathway and serotonin pathwayIn both pathways, the fi rst enzymatic step is the rate-limiting step. AADC: aromatic L-amino acid decarboxylase, AANAT: arylalkylamine N-acetyltransferase, AD: aldehyde dehydrogenase, 3HAO: 3-hydroxyanthranilic acid di-oxygenase, HIOMT: hydroxyindole-O-methyltransferase, IDO: indoleamine 2,3-dioxygenase, KAT: kynurenine ami-notransferase, KYN: kynureninase, MAO: monoamine oxidase, NAD+: nicotinamide adenine dinucleotide, QPRT: quinolinic acid phosphoribosyl transferase, TDO: tryptophan 2,3-dioxygenase, TPH: tryptophan hydroxylase, 5-HIAA: 5-hydroxyindoleacetic acid.

6

125


Supplementary Table S1. Overview of studies analyzing PD-L1 expression in NET Tumour/micro-environment characteristic

Population Method Percentage of positive cases in tumour, total cases

Percentage of positive cases in stroma or immune cells

First author and publication year

PD-L1, E1L3N NET patients who underwent resection

Tumor cells were positive if ≥ 10% tumor cells were positive, lymphocytes were positive if at least moderately staining was observed

15%, 48 35% intratumoural/ peritumoural cells

Cavalcanti 2017

PD-L1, unknown

Midgut NETs All cases with moderate or strong staining by Allred score were considered positive

69%, 32 Cives 2016

PD-L1, 9A11 Small intestine NETs

Based on published criteria

0%, 64 55% in stroma Da Silva 2016

PD-L1, 9A11 Pancreatic NET

Based on published criteria

11%, 18 17% in stroma Da Silva 2016

PD-L1, Ab28-8 Pulmonary NET

For each section, the approximate percentage of positive tumor cells and staining intensity determined the PD-L1 staining score

59%, 22 Fan 2016

PD-L1, 22C3 Resected carcinoids of the lung

Separately assessment of immune cells and tumor cells

0%, 57 0% in immune cells

Kasajima 2018

PD-L1, SP142 GEP-NEN Tumor cells were positive if ≥ 1% tumour cells were positive

0%, 15 Kim 2016

PD-L1, E1L3N Small intestine NETs

Specimens with ≥5% membranous expression were considered “positive”

13%, 70 24% in TILs Lamarca 2018

PD-L1, SP142 Pancreatic NETs

The staining was regarded as positive if its intensity on the membrane of the tumour cells was ≥ 2+ (on a scale of 0–3) and the percentage of positively stained cells was 5%

<5%, 75 Salem 2018

GEP-NEN; gastro-enteropancreatic neuroendocrine neoplasm, NET; neuroendocrine tumour, PD-L1; programmed death-ligand 1, TILs, tumour infiltrating lymphocytes.

7

Summary and future perspectives

7

129


Summary

Neuroendocrine tumors (NETs) are rare tumors with an incidence 3.5/100,000 per year and a prevalence of 35/100,000. The tumor arises from secretory cells of the neuroen-docrine system. Patients with NET may experience various symptoms from the tumor mass, the output of hormones secreted by the tumor, and treatment accompanying side effects. Patients often present with metastatic disease which cannot be cured. Patients with metastatic small intestine NET have a median survival of 100 months, for metastasized pancreatic NET median survival is 60 months (1). Compared to patients with other metastasized malignancies this is relatively long. In clinical trials with new drugs, the ideal endpoint is clear and objective, such as overall survival (OS) and quality of life (QoL). However, over the last years well-powered ran-domized controlled trials (RCTs) are performed in patients with NET, with progression free survival (PFS) as primary endpoint which serves as surrogate endpoint for OS. Given the heterogeneous patient population, the long natural course of the disease and the wide variability of subsequent lines of therapy after progression, the use of OS as primary endpoint is challenging in NET. Nevertheless, the aim of the research in this thesis was to measure and improve the clinical benefit of treatment and supportive care provided to patients with NET. We investigated tailored supportive measures for these patients as well as novel treatment approaches. One approach to improve QoL is with adequate information provision. Therefore, we investigated a web-based tailored information and support system. Furthermore, we studied if a dietary intervention and supplementation of deficient vitamins improved QoL. Moreover we analyzed whether le-sions with high [18F]FDOPA uptake could be detected, potentially to select these for local treatment. Finally, as immunotherapy is an upcoming treatment modality in oncology, we analyzed the NET immune microenvironment. There is currently limited knowledge of the composition of the immune microenvironment and the role of immunotherapy in patients with NET.In chapter 1 a general introduction and outline of the thesis is given. To define the clinical benefit of systemic antitumor treatment in patients with GEP-NET, in chapter 2 an overview of existing literature about trials comparing systemic antitu-mor treatment for GEP-NET versus a control is presented and their clinical benefit is determined. Papers were selected after search of the electronic databases (PubMed and Embase). Papers reporting comparative trials, conducted in adult patients with GEP-NET were included for analysis. The clinical benefit of the investigated therapies reported in the selected trials was determined according to the European Society of Medical Oncology Magnitude of Clinical Benefit Scale (ESMO-MCBS) and the net health benefit score according to the American Society of Clinical Oncology (ASCO) revised framework by four independent assessors (2-4). The ESMO-MCBS and the net

Chapter 7

130

health benefit (NHB) of ASCO revised framework could be applied, respectively, to only 7 and 9 trials investigating systemic treatments in patients with NET. The currently used systemic treatments for patients with GEP-NET had low NHB scores according to the revised ASCO framework and none could be graded as susbstantial clinical beneficial according to the ESMO-MCBS. QoL is an important factor that contributes to clinical benefit. Factors associated with enhanced QoL include; less clinical symptoms, and adequate information provision. Patients with a NET frequently experience physical and psychosocial complaints. Strategies to provide information to optimize supportive care for these patients are of interest. The internet was reported earlier as a useful source of general information for some patients with NET (5). In chapter 3 we aimed to examine whether a web-based tailored information and support system (WINS) reduces distress and/or improves patients’ perception and satisfaction of received information, QoL and empowerment. First a feasibility study was performed in newly diagnosed NET patients (chapter 3a). Participants were randomized between standard care (n = 10) or intervention with in addition to standard care, access to the WINS (n = 10) during 12 weeks. During stan-dard care all patients could consult an oncology nurse 24 hours a day, 7 days a week. Feasibility was measured by determination of the participation- and dropout rate. First of all, the use of the WINS was feasible. Furthermore contradictory effects on inform-ing and supporting patients with NET were obtained. Thereafter a properly powered clinical trial was performed. Patients with NET, stratified for those newly diagnosed (< 6 months) or a longer duration of the disease, were randomized between standard care and intervention, consisting of standard care and access to the WINS, during 12 weeks (chapter 3b). Patients completed questionnaires about distress with the distress thermometer and problem list, perception and satisfaction of received information with the QoL questionnaire (QLQ)-INFO 25, QoL with QLQ-C30 and QLQ-GINET21, and empowerment with the construct of empowerment (CEO) questionnaire. 92 Patients were randomized. The intervention group (n=53) also completed a questionnaire based on the technical acceptance model regarding their use of and opinion on the WINS. No difference in distress slope and slope of median global score on perceived information and satisfaction between the intervention and control group could be demonstrated. QoL did also not improve. Empowerment was better in the control group for 4 of the 5 constructs. We concluded after this trial that the WINS did not improved distress and patients’ perception and satisfaction of received information more in patients with NET compared to patients only receiving standard care. This illustrates that the information and support provided by an oncology nurse and medical doctor, during standard care remains an important source of information for patients with NET. Other important factors influencing QoL in patients with NET are gastro-intestinal prob-lems and increased frequency of bowel movements (6). These complaints could be

7

131


the consequence of tumor mass, secretion of neuroendocrine amines or somatostatin analogue (SSA) treatment. This can lead to impaired absorption of critical food compo-nents such as vitamin B12, the fat-soluble vitamins A, D, E, and K. Furthermore, in case of high serotonin production in NET patients, deficiency of tryptophan, the precursor of serotonin and niacin (vitamin B3), can lead to symptoms of niacin deficiency. With a feasibility study we aimed to explore if personalized dietary advice and supplementa-tion of deficient vitamins is feasible in patients with NET using a SSA. In this study, which is described in chapter 4, 15 patients with NET using a SSA for over 6 months were included. They were counseled by a dietician for a personalized dietary advice and got supplementation of deficient vitamins A, D, E, K, vitamin B12 and/or niacin. Feasibility was assessed by calculation of participation/dropout rate and number of (severe) adverse events related to the intervention. At baseline and after 4 and 18 weeks, concentrations of the fat-soluble vitamins A, D, E, and K and vitamin B12 were determined in plasma and the main metabolite of vitamin B3, N1-methylnicotinamide (N1-MN) was analyzed in 24-hour urine collections. QoL, distress, empowerment and nutrition state were assessed. The participation rate was 75% and dropout rate was 17%. No intervention related adverse events occurred. At start of the intervention, nine patients had one or more vitamin deficiencies. Despite the intervention including a personalized dietary advice and supplementation of deficient vitamins A, D, E, K, vitamin B12 or niacin, which improved vitamin levels in five patients, seven patients had one or more deficient vitamin levels after the intervention. New vitamin deficien-cies developed in three patients during study period. A heterogeneous effect on QoL, distress, empowerment and nutrition state was observed. A personalized dietary advice with vitamin supplementation of deficient vitamins seemed to be feasible in SSA-using patients with NET. Furthermore we demonstrated that SSA-using patients with NET are at risk for vitamin A, D, E, K, B12 and niacin deficiencies, which are not easily corrected by dietary advices or standard vitamin supplementation. Clinical symptoms determine, among other factors, QoL. In NET patients tumor mass and the secretion of neuroendocrine amines by tumor cells are the two main causes of symptoms. Selecting the most active amine producing tumor lesions for local treatment might be beneficial for patients with small intestinal (SI-) NET to reduce neuroendocrine symptoms, i.e. the carcinoid syndrome. Tumor burden, investigated via 18Fluorine-L-dihydroxyphenylalanine ([18F]FDOPA) uptake multiplied by volume was earlier shown to correlate with tumor markers of the serotonin and catecholamine pathway (7). There-fore, in chapter 5 we analyzed interlesional heterogeneity with [18F]FDOPA positron emission tomography (PET) scans in patients with SI-NET and investigated if lesions with substantially higher [18F]FDOPA uptake than most other lesions could be identified. Peak standard uptake value (SUV) of a lesion was divided by the SUV mean of the corresponding organ to determine [18F]FDOPA uptake. The magnitude of heterogeneity

Chapter 7

132

between lesions within a patient was calculated by dividing the lesion with highest by the one with lowest [18F]FDOPA uptake. Only in patients with ≥ 10 lesions, defined as evaluable patients, we determined whether lesions with high [18F]FDOPA uptake were present. Lesions with a higher tracer uptake than the upper inner or outer fence (> 1.5 or 3 times the interquartile range above the third quartile) were defined as lesions with mild or extreme high [18F]FDOPA uptake, respectively. In 38 patients, of whom 35 had a serotonin producing tumor, over 680 lesions were detected by [18F]FDOPA. The uptake of [18F]FDOPA varied with a median of 8-fold, up to 44-fold, between lesions within a patient. In 12 of 20 evaluable patients lesions with mild high [18F]FDOPA uptake were found and in five, lesions with extreme high [18F]FDOPA uptake. This leads to the conclusion that, [18F]FDOPA uptake shows considerable intrapatient heterogeneity. Furthermore in some patients one or more lesions with mild or extreme high [18F]FDOPA uptake could be detected. This suggests that in patients with a se-rotonin producing NET, [18F]FDOPA PET scans can support selection of lesions with substantially higher [18F]FDOPA uptake for local treatment. Chapter 6 describes a study with the aim to investigate the tumor immune microenvi-ronment in patients with serotonin and non-serotonin producing NETs. Immunotherapy with immune checkpoint inhibitors including programmed death (PD) and programmed death ligand-1 (PD-L1) antibodies have antitumor activity across numerous tumor types. Factors that are associated with clinical response on immune checkpoint inhibitors are presence of T-cells and deficiency of mismatch repair (MMR) proteins. The enzymes, indoleamine 2,3-dioxygenase (IDO) and tryptophan 2,3-dioxygenase (TDO) that deplete tryptophan in the tumor microenvironment along the kynurenine pathway, are also involved in immunity. The tryptophan depleted microenvironment leads to suppression of effector T-cells or the conversion to tumor tolerant regulatory T-cells, leading to cancer immunosuppression (8-10). In serotonin producing NETs (SP-NETs), IDO and TDO play an important role since tryptophan is also the precursor of the serotonin pathway. Little is known about the complex interactions of NET tumor cells with their surrounding immune microenvironment. Consequently knowledge about potential targets for immunotherapy in patients with NET is limited. Therefore, the aim of this study was to investigate the tumor immune microenvironment, i.e. the presence of PD-L1, T-cells, IDO, TDO, MMR proteins and cancer associated fibroblasts in tissue of treatment-naive SP-NETs and non-serotonin producing NET patients (NSP-NETs). Immunohistochemically analyses were performed for PD-L1, T-cells, IDO, TDO, MMR and activated fibroblasts. Tumors of 33 patients with SP-NET and 18 patients with NSP-NETs were studied. PD-L1 expression was seen in <1% of tumor and T-cells. T-cells were present in 33% of NETs, varying between 1-10% T-cells per high power field. IDO was expressed in tumor cells in 55% of SP-NETs and 22% of NSP-NETs (p=0.039). TDO was expressed in stromal cells in 64% of SP-NETs and 13% NSP-NETs (p=0.001).

7

133


None of the tumors had loss of MMR proteins. TDO positive stromal cells strongly expressed α-SMA and were therefore identified as cancer associated fibroblasts. We have demonstrated that NETs lack pre-existing immunity based on absence of PD-L1 expression by tumor or T-cells, presence of only few infiltrating T-cells and no MMR deficient tumor. This illustrates that two mechanisms are responsible for the cold im-mune microenvironment, namely the expression of IDO and TDO in a substantial part of NETs and the presence of cancer associated fibroblasts.

Future perspectivesIn our study with a web-based tailored information and support system, distress and satisfaction and perception of received information did not differ between the interven-tion group using the web based system and controls receiving standard of care. This is in contrast to another web-based system where patients could have a direct interac-tion with a nurse or physician. That randomized study was conducted in patients with advanced cancer. OS was prolonged in patients that were assigned to a program in which they could provide self-report symptoms at and between visits via a web-based questionnaire platform (11,12). In another web-based study, patients could monitor their symptoms, obtain individually tailored, evidence-based self-management support, ask questions to a clinical nurse specialist, communicate with other patients in a forum, and use a diary with the web-based system. The patients received questionnaires asking about reasons for using the different components of a specific WBS, and their usefulness (13). The highest rated component of that web-based system was e-mail communication with nurses, suggesting that more personalized information provision and communication is preferred. The contact and communication with health care professionals remain a crucial source of information and support for patients with NET. In the study reported in chapter 3, analyzing the WINS, patients in the control and intervention arm received standard care and therefore could call for every symptom or health question with a specialized nurse. The extensive standard care might be one of the reasons the WINS was not of additional value in patients with NET. Despite this, patients reported, they would like to obtain more information. These findings suggest that future web-based systems can have additional value in patients with NET if an interactive function is included that gives easy access to ask questions to a health care professional. We performed a feasibility study to improve symptoms that decrease QoL due to low vitamin status and unhealthy diet (chapter 4). With this study, we showed, in SSA-using patients with NET, that intervention with dietary advice and supplementation of defi-ciency of vitamin A, D, E, K, B12 and niacin is feasible. Despite the intervention, which improved vitamin levels measured in blood or urine almost half of the patients had one or more deficient vitamin level. A heterogeneous effect on QoL was observed. We are

Chapter 7

134

currently conducting a larger trial, (NCT03143946) in which we aim to decrease the number of NET patients with deficient vitamins A,D,E,K, B12 and niacin and thus to im-prove QoL. Based on the data obtained from the feasibility study, we increased the dose and duration of vitamin supplementation of vitamin D and vitamin K, respectively. Since we performed one of the first trials in SSA-using patients with NET addressing nutrition, still multiple research questions are left including; do patients develop deficiencies of other micronutrients, how can we prevent patients to develop vitamin deficiencies, what is the role of nutrition state in quality of life, do patients with NET have a higher chance to develop malnutrition, or in case of use of SSA, to develop hypoglycaemia? GEP-NET and its treatment could lead to changes in gastro-enteropancreatic anatomy, to pan-creas insufficiency and other toxicities and symptoms. Consequently, next to vitamin deficiencies, patients with NET probably also have higher risks to develop deficiencies of other micronutrients like, iron, calcium and folic acid. For future studies a stepwise approach is needed for SSA-using patients with NET, to determine how to improve the clinical benefit of a nutritional intervention. Currently there are not sufficient evidence based data on nutrition care for GEP-NET. Well-designed randomized trials are needed to get the knowledge in giving patients with NET optimal nutritional support.In patients with GEP-NET QoL can depend on high neuroendocrine amine excretion. For patients with refractory carcinoid syndrome on SSAs, loco regional therapy of lesions with high production of these amines could be considered (14). We showed that [18F]FDOPA uptake varied between lesions within a SI-NET patient (chapter 5). Furthermore in some patients one or more lesions with mild or extreme high [18F]FDOPA uptake could be detected. Potentially, in patients with carcinoid syndrome [18F]FDOPA PET scans can be used to select lesions with substantially higher [18F]FDOPA uptake for local treatment like radiofrequency ablation, transarterial (chemo-) embolization, resection or selective interval radiation therapy. Although prospective or comparative trials of these local treatment modalities are lacking, small series describe symptom-atic response rates of 75% for transarterial (chemo-) embolization, and an objective response rate of 50% for selective interval radiation therapy and transarterial (chemo-) embolization (15). In patients with colorectal liver metastases selective interval radia-tion therapy with Yttrium-90 added to standard chemotherapy had a better median PFS than chemotherapy without selective interval radiation therapy (16). Currently patients with neuroendocrine tumors with liver metastases are recruited for a trial analyzing treatment with selective interval radiation therapy with Holmium166 after PRRT treat-ment (NCT02067988). In another trial the potential benefit of local treatment with Yttrium90 labeled to 1,4,7,10-tetraazacyclododecane-N,N’,N″,N‴-tetraacetic acid-D-Phe1-Tyr3-octreotide (DOTATOC) is analyzed in NET patients with liver metastases (NCT03197012). Potentially, [18F]FDOPA PET scans can be of value for selecting patients with substantially higher [18F]FDOPA uptake for these local treatments.

7

135


Recently, a phase III trial assessed the efficacy and safety of telotristat ethyl and dem-onstrated a reduction in bowel movement frequency in the intervention group. This drug, in addition to SSA, was approved by the Food and Drug Administration and by the European Medicines Agency, for the treatment of patients with carcinoid syndrome diarrhea that SSA therapy has inadequately controlled (17). Unfortunately, this study did not investigate if the decreased bowel movements resulted in improved QoL. Improve-ment of disease-related symptoms could contribute to improvement in QoL. To improve insight in clinical benefit, measurement of QoL, must be reported in future trials. For NET patients with good prognosis, improvement of QoL is of major importance. Therefore, QoL measurement should be part of NET related trials. The RADIANT-4 trial investigated everolimus versus placebo in patients with advanced lung or gastro-intestinal NETs. Treatment with everolimus was associated with improvement in PFS which was the primary endpoint of this trial. No difference between the two groups was observed in time to deterioration in QoL which was one of the secondary endpoints. Two hypotheses were reported to explain these results (18). Experienced toxicity of evero-limus could be counterbalanced by the higher efficacy and potentially fewer disease related symptoms. Secondly, it was hypothesized that differences in QoL cannot be detected with the currently used methods. This might be a consequence of the indolent natural history of NET and the number of patients having symptoms at start of the study. Therefore, time to QoL deterioration would be a useful endpoint to measure in all NET trials, although not standardized (19). To improve ‘time to QoL deterioration’ as an endpoint, potential more simple methods to asses QoL in NET patients are needed, QoL assessments should be performed on a longer term, and a standardized approach to define time to deterioration is needed.In chapter 6 we have demonstrated that NETs lack pre-existing immunity as they do not express PD-L1 on tumor or T-cells, there are only few infiltrating T-cells and none of the tumors was MMR protein deficient. The expression of IDO and TDO in a substantial part of NETs and the presence of cancer associated fibroblasts suggest two mechanisms responsible for the cold immune microenvironment. IDO inhibitors are currently clinically tested for their role to restore immune function with or without PD-1 antibody inhibitors in phase II and III studies and an IDO1/TDO dual inhibitor is currently tested in a phase I study in patients with a solid tumor (NCT02073123 (20-23)). The phase III study, ECHO-301/KEYNOTE 252 was recently announced to be terminated early due to failure to improve PFS (24). In this study, patients with unresectable or metastatic melanoma were randomized to receive pembrolizumab in combination with either epacadostat (an IDO-inhibitor) or placebo. According to experts the IDO strategy has been moved to randomize clinical trials too fast (25). In our study we could not find an association between IDO expression and presence of T-cells and this might be an indication of the complex interaction of tumors and their immune microenvironment.

Chapter 7

136

Preclinical studies in mouse models of various cancer types showed that cancer associ-ated fibroblasts together with other stromal cells are being responsible for restricting the accumulation of T-cells in the vicinity of cancer cells (26). Furthermore, cancer associ-ated fibroblasts act by secreting various growth factors like transforming growth factor-β (TGF-β) (27,28). Increased TGF-β in the tumor immune microenvironment is recently recognized to represent a primary mechanism of immune evasion that promotes T-cell exclusion (29). Combining TGF-β blockade with immune checkpoint blockade in mouse models increases the anti-tumor efficacy of the therapy, suggesting that identifying and targeting micro-environmental regulators of anti-tumor immunity may increase the reach of immunotherapy approaches (30). In pancreatic NET cell lines, TGF-β leads to G1 growth arrest and the existence of an autocrine anti-proliferative loop was seen in neuroendocrine tumor cells (31). However the midgut carcinoid cell line, KRJ-I, lacking TGF-β growth inhibition, utilizes the TGF-β system to induce gene responses associated with growth promotion, invasion, and metastases, in contrast with normal, physiologic, enterochromaffin cell lines (32). In immunohistochemical analysis in 8 out of 12 NET samples expressed TGF-β and co-expression of TGF-β in both tumor and surrounding tissue was detected in 4 tumor samples (31). There is however limited information as yet with regards to the role of TGF-β in the NET immune microenvironment.Another potential drug, that could increase the immunity towards NETs and thereby their response to checkpoint inhibitors is interferon-α (IFN-α). Interferon is a registered drug for patients with NET (33). It plays a role in promoting tumor immunogenicity. IFN-α promotes a shift in host immunity against tumors, enhancing cell-mediated cytotoxicity and has a role in attracting CD4+ T-helper 1 lymphocytes traffic to the tumor (34). Furthermore CD8+ T-cells matured in the presence of IFN-α expressed significantly higher levels of PD-1 compared to CD8+ T-cells matured in presence of interleukin 12 (35). Combination treatment of IFN-α could potentially lead to higher levels of PD-1 and more targets to be treated by PD-1 therapy, and is therefore interesting to investigate. A phase I study in advanced melanoma analyzing pegylated-IFN with pembrolizumab in advanced melanoma has shown an acceptable toxicity profile and evidence of clinical efficacy (36). In another phase I study, the combination with pembrolizumab and IFN-α was explored in patients with advanced melanoma or renal cell carcinoma. Limited anti-tumor activity was observed. Response occurred in one of five patients with melanoma and two of 12 patients with renal cell carcinoma (37). Other clinical trials exploring the combination are ongoing. (NCT02112032, NCT02339324, (38)). Recently, the phase III trial SWOG SO518 analyzed the efficacy of bevacizumab plus octreotide LAR, com-pared with IFN-α plus octreotide LAR in patients with advanced and progressive NET. This study could not show an advantage of bevacizumab with a PFS of 16.6 above IFN-α 15.4 (HR 0.93 (CI95% 0.73-1.18)).

7

137


A long term control in tumor progression can be attained by IFN-α with an SSA in advanced NETs as presented in the SWOG SO518 (39,40). With the role of IFN-α in attracting CD4+ T-helper 1 lymphocytes traffic to the tumor, and leading to higher levels of PD-1 in T-cells, treatment with a PD-1 inhibitor combined with IFN-α probably become more effective in NET. Combining immune checkpoint blockade with micro-environmental regulators like TGF-β blockers or with IFN-α increases the anti-tumor efficacy of the therapy and is therefore promising for NETs, which lack pre-existing immunity. However the first clinical trials exploring these combination therapies are on-going in melanomas and other malignancies. To explore the effect of these combination therapies in patients with NET, the research has to be extended.

Chapter 7

138

References 1. Dasari A, Shen C, Halperin D, et al. Trends in the incidence, prevalence, and sur-

vival outcomes in patients with neuroendocrine tumors in the United States. JAMA Oncol 2017;3:1335-42.

2. Cherny NI, Sullivan R, Dafni U, et al. A standardised, generic, validated approach to stratify the magnitude of clinical benefit that can be anticipated from anti-cancer therapies: the Eu-ropean Society for Medical Oncology Magnitude of Clinical Benefit Scale (ESMO-MCBS). Ann Oncol 2015;26:1547-73.

3. Schnipper LE, Davidson NE, Wollins DS, et al. Updating the American Society of Clinical Oncology Value Framework: Revisions and reflections in response to comments received. J Clin Oncol 2016;34:2925-34.

4. Schnipper LE, Davidson NE, Wollins DS, et al. American Society of Clinical Oncology Statement: A conceptual framework to assess the value of cancer treatment options. J Clin Oncol 2015;33:2563-77.


6. Pearman TP, Beaumont JL, Cella D, Neary MP, Yao J. Health-related quality of life in pa-tients with neuroendocrine tumors: an investigation of treatment type, disease status, and symptom burden. Support Care Cancer 2016;24:3695-703.


8. Weinlich G, Murr C, Richardsen L, Winkler C, Fuchs D. Decreased serum tryptophan concentration predicts poor prognosis in malignant melanoma patients. Dermatology 2007;214:8-14.

9. Schrocksnadel K, Wirleitner B, Winkler C, Fuchs D. Monitoring tryptophan metabolism in chronic immune activation. Clin Chim Acta 2006;364:82-90.

10. Kema IP, Schellings AM, Hoppenbrouwers CJ, Rutgers HM, de Vries EG, Muskiet FA. High performance liquid chromatographic profiling of tryptophan and related indoles in body fluids and tissues of carcinoid patients. Clin Chim Acta 1993;221:143-58.

11. Basch E, Deal AM, Kris MG, et al. Symptom monitoring with patient-reported outcomes dur-ing routine cancer treatment: A randomized controlled trial. J Clin Oncol 2016;34:557-65.

12. Basch E, Deal AM, Dueck AC, et al. Overall survival results of a trial assessing patient-reported outcomes for symptom monitoring during routine cancer treatment. JAMA 2017;318:197-8.

13. Ruland CM, Andersen T, Jeneson A, et al. Effects of an internet support system to assist cancer patients in reducing symptom distress: a randomized controlled trial. Cancer Nurs 2013;36:6-17.

14. Pavel M, O‘Toole D, Costa F, et al. ENETS Consensus Guidelines Update for the manage-ment of distant metastatic disease of intestinal, pancreatic, bronchial neuroendocrine neo-plasms (NEN) and NEN of unknown primary site. Neuroendocrinology 2016;103:172-85.

15. de Mestier L, Zappa M, Hentic O, Vilgrain V, Ruszniewski P. Liver transarterial emboliza-tions in metastatic neuroendocrine tumors. Rev Endocr Metab Disord 2017;18:459-71.

16. van Hazel GA, Heinemann V, Sharma NK, et al. SIRFLOX: Randomized phase III trial comparing first-line mFOLFOX6 (plus or minus bevacizumab) versus mFOLFOX6 (plus

7

139


or minus bevacizumab) plus selective internal radiation therapy in patients with metastatic colorectal cancer. J Clin Oncol 2016;34:1723-31.

17. Kulke MH, Horsch D, Caplin ME, et al. Telotristat ethyl, a tryptophan hydroxylase inhibitor for the treatment of carcinoid syndrome. J Clin Oncol 2017;35:14-23.

18. Walter T. Maintaining quality of life for patients with neuroendocrine tumours. Lancet Oncol 2017;18:1299-300.

19. Anota A, Hamidou Z, Paget-Bailly S, et al. Time to health-related quality of life score de-terioration as a modality of longitudinal analysis for health-related quality of life studies in oncology: do we need RECIST for quality of life to achieve standardization? Qual Life Res 2015;24:5-18.

20. Pilotte L, Larrieu P, Stroobant V, et al. Reversal of tumoral immune resistance by inhibition of tryptophan 2,3-dioxygenase. Proc Natl Acad Sci U S A 2012;109:2497-502.

21. Mellemgaard A, Engel-Norregaard L, Andersen MH, Zocca M-, Svane IM. Combination immunotherapy with IDO vaccine and PD-1 inhibitors in advanced NSCLC. J Clin Oncol 2017;35.

22. Brochez L, Chevolet I, Kruse V. The rationale of indoleamine 2,3-dioxygenase inhibition for cancer therapy. Eur J Cancer 2017;76:167-82.

23. Prendergast GC, Malachowski WP, DuHadaway JB, Muller AJ. Discovery of IDO1 inhibi-tors: From bench to bedside. Cancer Res 2017;77:6795-811.

24. Incyte and Merck provide update on phase 3 study of epacadostat in combination with KEYTRUDA® (pembrolizumab) in patients with unresectable or metastatic melanoma. Business Wire. (2018, at https://www.businesswire.com/news/home/20180406005141/en/Incyte-Merck-Provide-Update-Phase-3-Study).

25. K G. A promising new cancer drug has hit a major setback, raising questions about whether the field is moving too fast. 2018;.

26. Joyce JA, Fearon DT. T cell exclusion, immune privilege, and the tumor microenvironment. Science 2015;348:74-80.

27. Kakarla S, Song XT, Gottschalk S. Cancer-associated fibroblasts as targets for immuno-therapy. Immunotherapy 2012;4:1129-38.

28. Kalluri R. The biology and function of fibroblasts in cancer. Nat Rev Cancer 2016;16:582-98.

29. Mariathasan S, Turley SJ, Nickles D, et al. TGFbeta attenuates tumour response to PD-L1 blockade by contributing to exclusion of T cells. Nature 2018;554:544-8.

30. Tauriello DVF, Palomo-Ponce S, Stork D, et al. TGFbeta drives immune evasion in geneti-cally reconstituted colon cancer metastasis. Nature 2018;554:538-43.

31. Wimmel A, Wiedenmann B, Rosewicz S. Autocrine growth inhibition by transforming growth factor beta-1 (TGFbeta-1) in human neuroendocrine tumour cells. Gut 2003;52:1308-16.

32. Kidd M, Schimmack S, Lawrence B, Alaimo D, Modlin IM. EGFR/TGFalpha and TGFbeta/CTGF signaling in neuroendocrine neoplasia: theoretical therapeutic targets. Neuroendo-crinology 2013;97:35-44.

33. Indications for interferon alfa 2018. European Medicines Agency. (at http://www.ema.europa.eu/ema/index.jsp?curl=pages/medicines/human/medicines/000281/human_med_000846.jsp&mid=WC0b01ac058001d124»).

34. Rafique I, Kirkwood JM, Tarhini AA. Immune checkpoint blockade and interferon-alpha in melanoma. Semin Oncol 2015;42:436-47.

Chapter 7

140

35. Gerner MY, Heltemes-Harris LM, Fife BT, Mescher MF. Cutting edge: IL-12 and type I IFN differentially program CD8 T cells for programmed death 1 re-expression levels and tumor control. J Immunol 2013;191:1011-5.

36. Davar D, Wang H, Chauvin J-, et al. Phase IB study of pembrolizumab (Pembro) and pegylatedinterferon alfa-2b (Peg-IFN) in advanced melanoma (MEL). J Clin Oncol 2016;34:abstract.

37. Atkins MB, Hodi FS, Thompson JA, et al. Pembrolizumab plus pegylated interferon alfa-2b or ipilimumab for advanced melanoma or renal cell carcinoma: Dose-finding results from the phase Ib KEYNOTE-029 study. Clin Cancer Res 2018;24:1805-15.

38. He AR, Morse M, Iyer RV, et al. A phase II multicenter study evaluating combination im-munotherapy with pembrolizumab and peginterferon alfa-2b for advanced cholangiocarci-noma. J Clin Oncol 2017;35:abstract.

39. Amoroso V, Pavel M, Claps M, et al. IFN-alpha in advanced well-differentiated neuroendo-crine tumors: the neglected drug? Future Oncol 2018;14:897-9.

40. Yao JC, Guthrie KA, Moran C, et al. Phase III prospective randomized comparison trial of depot octreotide plus interferon alfa-2b versus depot octreotide plus bevacizumab in patients with advanced carcinoid tumors: SWOG S0518. J Clin Oncol 2017;35:1695-703.

Nederlandse samenvatting

145


Neuro-endocriene tumoren (NETs) zijn zeldzame tumoren met een incidentie van 3,5/100.000 per jaar en een prevalentie van 35/100.000. De tumor ontstaat vanuit cel-len van het neuro-endocriene systeem. Patiënten met een NET kunnen verscheidene symptomen ervaren bijvoorbeeld door de massawerking en/of afgifte van hormonen door de tumor, of door bijwerkingen gerelateerd aan de behandeling. Bij presentatie is de ziekte vaak al gemetastaseerd en kunnen patiënten meestal niet meer genezen worden. Patiënten met een gemetastaseerde dunne darm NET hebben een mediane overleving van 100 maanden, voor patiënten met een gemetastaseerd pancreas NET is dat 60 maanden. In klinische studies waarin nieuwe geneesmiddelen worden getest is het ideale eind-punt helder en objectief, bijvoorbeeld de duur van overleving en kwaliteit van leven (KvL). Toch zijn de laatste jaren, grote gerandomiseerde studies bij patiënten met NET gedaan, waarin progressie vrije overleving als primair eindpunt werd gebruikt, als surro-gaat eindpunt voor overleving. Bij NET is het gebruik van overleving als primair eindpunt moeilijk omdat de patiënten populatie heterogeen is, het ziektebeloop langdurig en er meerdere lijnen behandeling mogelijk zijn. Het doel van het onderzoek beschreven in dit proefschrift was om de behandeling en zorg aan patiënten met een NET te meten en te verbeteren. We onderzochten bij deze patiënten individuele ondersteunende maatregelen en nieuwe manieren van behandelen. Ten eerste richtten we ons op de verbetering van KvL van patiënten met een NET. We onderzochten een online informatie- en ondersteuningssysteem op maat en een inter-ventie met dieet adviezen en suppletie van deficiënte vitamines. . Daarnaast hebben we gekeken of we op 18Fluorine-L-dihydroxyphenylalanine (18F-DOPA) PET scans van patiënten met een NET laesies met hoge opname konden worden geïdentificeerd, met als doel om deze te selecteren voor lokale behandelingen. Ten slotte, hebben we de tumor en zijn immunologische micromilieu onderzocht. Bij andere maligniteiten blijkt dit immunologische micromilieu van belang voor het effect van immunotherapie. Bij NETs is nog weinig bekend over de rol die immunotherapie zou kunnen spelen in de behandeling. In hoofdstuk 1 is een algemene introductie en uiteenzetting van het proefschrift be-schreven. Om het klinische voordeel van systemische anti-tumor behandeling in patiënten met een NET van gastro-intestinale of pancreas origine (GEP-NET) in kaart te brengen, is in hoofdstuk 2 een overzicht gegeven van bestaande literatuur over studies die anti-tumor behandeling voor GEP-NET vergelijken met een controle behandeling of placebo, en is het klinische nut ervan bepaald. Artikelen werden geselecteerd aan de hand van een zoekopdracht in elektronische databases (PubMed en Embase). Artikelen die vergelij-kende studies beschreven, uitgevoerd in volwassen patiënten met GEP-NET werden geïncludeerd voor analyse. Het klinisch nut van de onderzochte behandelingen in de

146


geselecteerde studies werd bepaald aan de hand van de European Society of Medical Oncology Magnitude of Clinical Benefit Scale (ESMO-MCBS) en de ‘net health benefit score’ (NHB) van het gereviseerde ‘American Society of Clinical Oncology (ASCO) framework’ door 4 onafhankelijke Nederlandse beoordelaars. De ESMO-MCBS kon op slechts zeven studies, die behandelingen in patiënten met NET onderzochten, worden toegepast en de NHB van gereviseerde ASCO framework op negen studies. Geen en-kele behandeling die momenteel wordt gebruikt voor patiënten met GEP-NET voldeed aan de door de ESMO-MCBS gestelde criteria voor substantieel klinisch nut, en deze behandelingen hadden lage NHB scores volgens het gereviseerde ASCO framework. Ondanks de lage incidentie, de heterogene patiëntenpopulatie en het relatief lange ziekte beloop, zouden toekomstige studies die nieuwe behandelingen onderzoeken, zich moeten richten op het verbeteren van eindpunten als overleving en kwaliteit van leven.Het verbeteren van de KvL van patiënten met NET is een belangrijk onderdeel van een klinisch nuttige behandeling. Factoren die geassocieerd worden met een betere KvL, zijn het hebben van minder klinische symptomen en het krijgen van adequate informatie. Patiënten met een NET ervaren vaak fysieke en psychosociale klachten. Om optimale ondersteunende behandeling voor deze patiënten te kunnen geven, zijn er verschil-lende methoden. Het internet bleek een bruikbare informatiebron voor algemene infor-matie voor sommige patiënten met een NET. In hoofdstuk 3 hebben we onderzocht of een online informatie- en ondersteuningssysteem (WINS) klachten vermindert en/of de perceptie en tevredenheid over ontvangen informatie, KvL en empowerment verbetert. Eerst werd een haalbaarheidsstudie verricht in patiënten met een recent gediagnosti-ceerde NET (hoofdstuk 3a). Deelnemers werden gerandomiseerd tussen standaard-behandeling en standaardbehandeling met interventie door middel van toegang tot de WINS (n=10) gedurende 12 weken. Gedurende de standaardbehandeling konden patiënten laagdrempelig vragen stellen aan een arts of oncologieverpleegkundige op elk moment en op elke dag. De haalbaarheid werd bepaald aan de hand van het aantal patiënten dat aan de studie deelnam en uitviel. Gebruik van de WINS bleek haalbaar. Er werden tegenstrijdige effecten op perceptie en tevredenheid over verkregen informatie gevonden. Aan de hand van deze informatie, werd een voldoende gepowerde studie opgezet. Patiënten met een NET, gestratificeerd voor degenen met een recent gestelde diagnose (<6 maanden) en degenen met een langere ziekteduur werden gerandomi-seerd tussen standaardbehandeling met of zonder toegang tot de WINS, gedurende 12 weken (hoofdstuk 3b). Patiënten vulden vragenlijsten in over hun klachten door middel van de lastmeter en een probleemlijst. Met de KvL vragenlijst (QLQ)-INFO 25 mat de perceptie en tevredenheid over verkregen informatie. KvL werd met de QLQ-C30 en QLQ-GINET21 gemeten en empowerment met de samengestelde empowerment vra-genlijst. 92 Patiënten werden gerandomiseerd. De interventie groep (n=53) vulde ook

147


een vragenlijst ten aanzien van hun gebruik en mening over de WINS. De vragenlijst was op een model gebaseerd dat gebruikt wordt om de individuele acceptatie en het gebruik van websites uit te leggen. Er werd geen verschil in afname van klachten of in toename van tevredenheid over ontvangen informatie gemeten, ook werd de KvL niet beter in de interventie groep. Empowerment was beter in de controlegroep voor 4 van de 5 onderdelen. Uit deze studie concludeerden we dat de WINS in patiënten met een NET zowel de klachten als de tevredenheid over ontvangen informatie niet verbeterde, in vergelijking met patiënten die alleen de standaardbehandeling ontvingen. Dit toont dat de informatie en ondersteuning die gegeven wordt door een oncologieverpleegkun-dige en arts tijdens standaardbehandeling een zeer belangrijke bron van informatie is voor patiënten met een NET.Andere belangrijke factoren die de KvL in patiënten met een NET beïnvloeden zijn gastro-intestinale klachten, waaronder het hebben van frequente ontlasting. Deze klachten kunnen veroorzaakt worden door tumormassa, uitgescheiden neuro-endocriene aminen of door behandeling met somatostatine analogen (SSA’s). Dit kan leiden tot verminderde absorptie van essentiële nutriënten zoals vitamine B12 en de vet-oplosbare vitamines A, D, E, en K. Bovendien, kunnen bij patiënten met een hoge serotonineproductie symptomen passend bij vitamine B3 deficiëntie ontstaan, via defi-ciëntie van tryptofaan, wat een voorloper is van zowel serotonine, als niacine (vitamine B3). Met een haalbaarheidsstudie wilden we beoordelen of een persoonlijk dieet en suppletie van deficiënte vitamines haalbaar is in patiënten met een NET die een SSA gebruiken. In deze studie, welke is beschreven in hoofdstuk 4, zijn 15 patiënten met een NET geïncludeerd, die langer dan 6 maanden een SSA gebruikten. Zij werden gezien door een diëtist die ze een persoonlijk dieetadvies gaf en deficiënte vitamines A, D, E, K, B12 en/of nicotinamide werden gesuppleerd. De haalbaarheid werd bepaald aan de hand van het aantal patiënten wat de studie deelnam en uitviel en door het bijhouden van het aantal (ernstige) bijwerkingen gerelateerd aan de interventie. Bij start, na 4 en 18 weken werden de waarden van vitamine A, D, E, K en B12 bepaald in plasma en de belangrijkste metaboliet van vitamine B3, N1-methylnicotinamide (N1-MN) werd gemeten in 24-uurs urine. KvL, klachten, empowerment en voedingsstatus werden gemeten. Vijfenzeventig procent van de patiënten namen deel, 17% viel uit. Er traden geen bijwerkingen op gerelateerd aan de interventie. Vooraf aan de interventie hadden negen patiënten één of meer vitaminedeficiënties. Ondanks de interventie, welke in vijf patiënten vitaminewaardes verbeterde, hadden zeven patiënten na de interventie nog één of meer vitaminedeficiënties. In drie patiënten waren nieuwe vitaminedeficiënties ontstaan tijdens de studie. Een heterogeen effect op KvL, klachten, empowerment en voedingsstatus werd gezien. Een persoonlijk dieetadvies met suppletie van vita-minedeficiënties bleek haalbaar in patiënten met NET, die een SSA gebruiken. Verder toonden we dat deze patiënten risico lopen om vitamine A, D, E, K, B12 en vitamine B3

148


deficiënties te ontwikkelen, die niet makkelijk te corrigeren zijn door dieet adviezen en standaard vitaminesuppletie.Vermindering van symptomen zijn een van de factoren die bijdragen aan toename van KvL. In patiënten met een NET zijn tumormassa en de secretie van neuro-endocriene aminen door de tumor de twee belangrijkste oorzaken voor symptomen. Selectie van de meest actieve laesies voor lokale behandeling kan zorgen voor afname van neuro-endocrine klachten bij patiënten met een dunne darm NET. Tumorlast, bepaald via 18F-DOPA opname vermenigvuldigd met volume correleert met activiteit van de serotonine en catecholamine route. Door analyse van 18F-DOPA PET scans werd interlesionale heterogeniciteit in patiënten met dunne darm NET onderzocht en werd beoordeeld of laesies met substantieel hogere 18F-DOPA activiteit ten opzichte van de meeste andere laesies konden worden geïdentificeerd. Deze studie is in hoofdstuk 5 beschreven. Daarin zijn 18F-DOPA PET scans geanalyseerd. Om de 18F-DOPA op-name voor elke tumorlaesie te bepalen werd de piek standard uptake value (SUV) van de laesie gedeeld door de gemiddelde SUV van het bijpassende orgaan. De grootte van de heterogeniciteit tussen laesies binnen een patiënte werd berekend door de laesies met de hoogste 18F-DOPA opname te delen door de laesie met de laagste opname. Dit hebben we geanalyseerd bij patiënten met ≥ 10 laesies, gedefinieerd als te evalueren patiënten. We hebben bepaald of laesies met hoge 18F-DOPA opname aanwezig wa-ren, in vergelijking met andere laesies in die patiënt. In 38 patiënten, waarvan 35 met een serotonine-producerende tumor, werden meer dan 680 laesies gedetecteerd met behulp van de 18F-DOPA PET scan. De 18F-DOPA opname varieerde met een mediaan van 8 keer, tot aan 44-keer tussen laesies binnen één patiënt. In 12 van 20 patiënten met ≥ 10 laesies werden laesies met mild verhoogde 18F-DOPA opname gezien en vijf patiënten hadden laesies met extreem verhoogde 18F-DOPA opname. Dit toonde dat 18F-DOPA tumor opname aanzienlijk heterogeen is binnen deze patiënten. Verder konden in sommige patiënten één of meer laesies met substantieel hogere 18F-DOPA opname gedetecteerd worden. Dit suggereert dat in patiënten met een serotonine-producerende NET, 18F-DOPA PET scans gebruikt zouden kunnen worden om laesies te selecteren met een substantieel hogere 18F-DOPA opname voor lokale behandeling. In hoofdstuk 6 wordt een studie beschreven waarin het immunologische micromilieu van serotonine-producerende en niet-serotonine-producerende NETs is onderzocht. Immunotherapie met immuun checkpoint remmers, zoals ‘programmed death’ (PD1) en ‘programmed death ligand-1’ (PD-L1) antilichamen, laten antitumor activiteit zien tegen vele tumorsoorten. De respons kans op immuun checkpoint remmers wordt geassocieerd met meerdere factoren, zoals aanwezigheid van T-cellen en afwezigheid van mismatch repair (MMR) eiwitten. Ook de enzymen, indoleamine 2,3-dioxygenase (IDO) en tryptofaan 2,3-dioxygenase (TDO), die tryptofaan verlagen in het tumor micromilieu via de kynurenine route, zijn betrokken bij immunosuppressie. Het tumor

149


micromilieu met laag tryptofaan leidt tot vermindering van effector T-cellen en omzetting naar regulatoire T-cellen, wat leidt tot remming van de immuunrespons gericht tegen de tumorcellen. Doordat tryptofaan, dat onder andere door IDO en TDO wordt omgezet, ook de voorloper is in de serotonine route, hebben patiënten met een serotonine-producerende NET vaak een lagere tryptofaanspiegel en is de rol van deze enzymen bij deze tumoren nog interessanter. Er is weinig bekend over de complexe interacties van NET-cellen met hun immunologische micromilieu. Dit betekent ook dat de kennis over eventuele aangrijpingspunten voor immuun checkpoint remmers in NET beperkt is. Het doel van deze studie was om het immunologische tumor micromilieu, dat wil zeggen de aanwezigheid van PD-L1, T-cellen, IDO, TDO, MMR eiwitten en kanker geassocieerde fibroblasten in weefsel van patiënten met serotonine-producerende en niet-serotonine-producerende NET vooraf aan behandeling te onderzoeken. Immunohistochemische analyses werden uitgevoerd voor PD-L1, T-cellen, IDO, TDO, MMR-eiwitten en ge-activeerde fibroblasten. Tumoren van 33 patiënten met serotonine-producerende NET en 18 patiënten met niet-serotonine-producerende NETs werden beoordeeld. PD-L1 expressie werd gezien in <1% van de tumor en T-cellen. T-cellen, waren aanwezig in 33% van de NETs. In de NETs waar T-cellen aanwezig waren, werden 1-10% T-cellen per high power field gezien. IDO-expressie was aanwezig in 55% van de tumorcellen van serotonine-producerende NETs en 22% van niet-serotonine-producerende NETs (p=0,039). TDO was aanwezig in stromacellen in 64% van de tumoren van patiënten met serotonine-producerende NET en in 13% van niet-serotonine-producerende NET (p=0,001). Geen van de tumoren had verlies van MMR-eiwitten. Stromacellen met TDO-expressie hadden ook sterke expressie van α-SMA en werden daarom als kanker geassocieerde fibroblasten geduid. In deze studie hebben we getoond dat het tumor micromilieu weinig immunogeen is. Door de IDO- en TDO- expressie in een aanzienlijk deel van de NETs en de aanwezigheid van kanker geassocieerde fibroblasten zouden er twee mechanismen verantwoordelijk kunnen zijn voor dit weinig actieve immunologi-sche micromilieu. Verdere karakterisering van het tumor micromilieu zal inzicht geven in wat de eventuele mogelijkheden zijn voor rationele immunotherapie voor patiënten met een NET.

Dankwoord

153

Dankwoord

Met veel plezier heb ik afgelopen jaren aan dit proefschrift gewerkt, ik ben dan ook zeer verheugd dat het gelukt is om dit project tot een goed einde te brengen. Het was fijn om te merken dat velen me daarin, soms op verschillende manieren, wilden ondersteunen. Veel dank gaat uit naar alle mensen die hebben geholpen bij de totstandkoming van dit proefschrift. Een aantal wil ik graag in het bijzonder noemen:

Allereerst de patiënten; hartelijk dank voor jullie deelname aan één of meer van de onderzoeken. Sommigen van jullie namen binnen een (paar) jaar wel aan drie onder-zoeken deel.

Dr A.M.E. Walenkamp, beste Annemiek, vriendelijk, laagdrempelig te benaderen, en positief kritisch; Van jou heb ik geleerd om gestructureerd te denken en dat te gebruiken voor het schrijven van een goed artikel. Ik kon altijd bij je binnen lopen voor een vraag, een oplossing voor een probleem of een handtekening. Zodat ik daarna vol positieve energie en letterlijk opgewarmd weer kon doorwerken. De eerste paar maanden heb ik even moeten wennen, maar ik vind en vond je begeleiding erg fijn en heb me altijd serieus genomen en gesteund gevoeld en doe dat nog. Heel hartelijk dank! Graag zet ik onze samenwerking zo voort.

Prof dr E.G.E. de Vries, beste Liesbeth, graag wil ik je hartelijk bedanken. Bij onze besprekingen en in het verbeteren van conceptversies van manuscripten, heb ik veel gehad aan je analytische blik, jij keek naar de verbanden met oncologie in bredere zin, kon goed ‘out of the box’ denken en bracht een project vaak verder. Ook je geweldige benadering naar patiënten toe en enthousiasme, hebben me veel geleerd. Je uitspraak bij de soms volle poli’s zal ik niet snel vergeten; ‘We trekken gewoon de trimschoenen aan.’

Beste leden van de manuscriptcommissie, professor de Herder, professor Hollema en professor Hospers, veel dank voor de beoordeling van dit proefschrift.

Verder wil ik graag alle coauteurs bedanken voor hun kritische en vaak specialistische inbreng;

Grytsje, hartelijk dank, we hebben goed samen gewerkt en het was ook heel fijn dat ik van jóú de taken mocht overnemen. Bij overname waren de data al piekfijn op orde. Bij vervolgstudies probeerde ik deze georganiseerde manier van werken zoveel mogelijk mee te nemen.

154

Dankwoord

Dr Kats-Ugurlu, beste Gursah, urenlang hebben we achter de microscoop de details van NET tumoren bestudeerd, om dan te bedenken dat we toch ook nog een andere immunohistochemische kleuring wilden toepassen op de coupes. Tussendoor had je gelukkig wel nog de mogelijkheid om mij adviezen over dit promotietraject te geven. Hartelijk dank voor je engelengeduld en behulpzaamheid.

Prof dr Kema, en Martijn van Faassen, onze overleggen waren altijd inspirerend; de invloed van bananen op de samenstelling van urine, of molecuulformules van het indolprofiel, na elk overleg was ik weer een stukje wijzer. Hartelijk bedankt voor deze samenwerking.

Prof dr Brouwers en prof dr Boellaard, beste Adrienne en Ronald, hartelijk voor jullie begeleiding bij het project waarbij door middel van [18F]FDOPA PET scans NET laesies analyseerden. De theorie achter de moleculaire beeldvorming vond ik erg interessant.

Henny, het was erg goed om met je samen te werken; je nuchtere houding, je mensen-kennis, het vertrouwen dat iedereen in je had, van professor tot patiënt, is bewonde-renswaardig! Vaak konden patiënten nog even naar jou voor verdere uitleg of support, wat zij en ik zeer waardeerden.

Sietske en Bea; heel fijn dat jullie het ‘DIVIT-project’ met ons aan wilden gaan. Het heeft veel tijd gekost, maar we hebben door middel van de pilotstudie wel de aandacht kunnen geven aan voeding en vitamines bij patiënten met NET, een onderwerp waar bijna niks over bekend was. Met de gegevens daaruit konden we de vervolgstudie in uitgebreidere vorm opzetten.

Alle medewerkers bij de afdeling Medische Oncologie, die ik nog niet genoemd heb, hartelijk dank voor alle medewerking en de goede sfeer.

‘Roomies’ en U4; Kirsten, Iris, Maaike, Lars, Pieter, Thijs, Clarieke, Suzanne, Niek, Kees, Sjoukje, Frederike, Johannes, Harm, de anderen en de 3 echte ‘guppen’! Het was ‘n geweldige tijd! Jullie waren geweldige collega’s, altijd in voor een grap, gesprek of luisterend oor.

Lieve vriendinnen, lieve Marlies, hartelijk dank voor alle keren dat je heerlijk voor me hebt gekookt, een rondje mee ging hardlopen en de kopjes (of bierpul met) thee waarbij we alles bespraken.

155

Dankwoord

Beste Inge, Loes, Annelies, Kim, Tanja, Floor en Marlieke wat hebben we gezellig gewandeld, weekendjes gevierd en gekletst. Hopelijk kan dat nu weer vaker. Heel leuk om te zien hoe jullie gezinnen uitbreidden en jullie daarin in je nieuwe rol groeiden. Heel erg bedankt!

Beste loopmaatjes, Simon, Marlies, Egbert, GA LA1, LA2 en trainers; na een dag achter mijn bureau, was t heerlijk om met jullie de benen te strekken, uit te waaien, en al inspannend te ontspannen, al of niet gevolgd door een goed resultaat, of koffie en taart. Dank jullie wel.

Beste paranimfen; Lieve Iris, ik leerde je pas later in het traject kennen, maar ondanks je harde werken, voelde je vaak goed aan wat er speelde. Het was fijn om tot en met ‘t eind een maatje op Y3 te weten en te hebben. Hartelijk dank. Lieve Kim, heel fijn dat je nu naast me staat. Mijn zusje, met wie ik als kind monopoly speelde, die het snapt als ik flink wil sporten en die ik altijd kan bellen. Hartelijk dank.

Lieve familie, pap, mam, Maart & Rich, Gijs en Kim, Karen en Marielle, en natuurlijk mijn gezellige neefjes en nichtjes. Heel erg bedankt voor al jullie steun, direct en indirect, door soms voor afleiding te zorgen met een relaxed weekendje in Neerloon of met ‘de kids’, of juist door mee te kijken naar een mailtje of grammaticale zin.

Allen; hartelijk bedankt voor alles!

Date post:	31-Jul-2020
Category:	Documents
Upload:	others
View:	5 times
Download:	0 times

University of Groningen Neuroendocrine tumors; measures to ...Lotte D. de Hosson. Neuroendocrine...

Documents